Ansa Evidence Console

Repo-owned dashboard snapshot from existing evidence only. Unknowns stay visible; this page does not infer usage, token cost, maturity, success rate, or route chains.

Readiness2 met / 3 partial / 1 insufficient
Unknowns1 readiness KPIs
Routes surfaced69 across 8 realms
Components366 registry-backed entries
Snapshot2026-07-09T22:10:00.000Z
Source commitunknown
Generator sourcescripts/core/generate-evidence-console.mjs
Generator versionnot_available
Canonical URLhttps://ansa.eunoetic.com/
Freshnessnot_stale (0 days)
not_stale Snapshot timestamp 2026-07-09T22:10:00.000Z; freshness checked 2026-07-09T22:10:00.000Z; canonical URL https://ansa.eunoetic.com/; deployment identity unknown.
Active routes69fact
Active skills12fact
Active workflows3fact
Active templates90fact
Measured PRs86fact
Measured receipts112fact
Validation commands637fact
Usage and cost evidenceobserved usage; exact totals incompleteobserved usage; exact totals incomplete

Component Catalogue

Component Taxonomy

Type Meaning Rows
adapter Repo or tool boundary guidance from adapter registry entries. 33
agent Codex staged worker role surfaced from the staged execution profile. 14
generated_skill Installed Codex skill wrapper generated from active Ansa registry metadata. 46
prompt Operator or route prompt/template entry from the prompt registry. 65
route Active route registry entry that points at an entrypoint workflow or surface. 69
skill Active shared Ansa skill registry entry. 12
template Active reusable template entry declared in the template registry. 90
workflow Workflow entrypoint declared in the workflow registry. 3
script Repo script entry declared in the script registry. 34

Why Unknown

unknown
No source-backed value is present in the active catalogue, registry, or generated evidence.
not_available
The field is not applicable to this component type or no deterministic invocation/source exists.
not yet measured
The later #613/#618 substrate is expected to provide evidence; this dashboard does not infer measurements.
adapter-cost-policy adapter active not_available not_available unknown not yet measured not_available
adapters-root-contract adapter active not_available not_available unknown not yet measured not_available
ansa-repo-adapter adapter active not_available not_available unknown not yet measured not_available
brain2-repo-adapter adapter active not_available not_available unknown not yet measured not_available
cartant-operator-pack-repo-adapter adapter active not_available not_available unknown not yet measured not_available
cartant-tool-adapter adapter active not_available not_available unknown not yet measured not_available
chatgpt-tool-adapter adapter active not_available not_available unknown not yet measured not_available
claude-code-tool-adapter adapter active not_available not_available unknown not yet measured not_available
coderabbit-review-participant-adapter adapter active not_available not_available unknown not yet measured not_available
codex-code-assist-review-participant-adapter adapter active not_available not_available unknown not yet measured not_available
codex-runner-notes adapter active not_available not_available unknown not yet measured not_available
codex-staged-execution-profile adapter active not_available not_available unknown not yet measured not_available
codex-tool-adapter adapter active not_available not_available unknown not yet measured not_available
copilot-code-review-participant-adapter adapter active not_available not_available unknown not yet measured not_available
copilot-tool-adapter adapter active not_available not_available unknown not yet measured not_available
cordis-repo-adapter adapter active not_available not_available unknown not yet measured not_available
enki-repo-adapter adapter active not_available not_available unknown not yet measured not_available
gemini-code-assist-review-participant-adapter adapter active not_available not_available unknown not yet measured not_available
gemini-runner-notes adapter active not_available not_available unknown not yet measured not_available
gemini-tool-adapter adapter active not_available not_available unknown not yet measured not_available
goose-tool-adapter adapter active not_available not_available unknown not yet measured not_available
imelda-repo-adapter adapter active not_available not_available unknown not yet measured not_available
litellm-openrouter-gateway adapter active not_available not_available unknown not yet measured not_available
opencode-runner-notes adapter active not_available not_available unknown not yet measured not_available
opencode-staged-model-profile adapter active not_available not_available unknown not yet measured not_available
opencode-tool-adapter adapter active not_available not_available unknown not yet measured not_available
repo-adapter-contract adapter active not_available not_available unknown not yet measured not_available
repo-adapter-template adapter active not_available not_available unknown not yet measured not_available
review-participant-adapter-contract adapter active not_available not_available unknown not yet measured not_available
stage-model-routing-standard adapter active not_available not_available unknown not yet measured not_available
tool-adapter-template adapter active not_available not_available unknown not yet measured not_available
traderpulse-example-repo-adapter adapter active not_available not_available unknown not yet measured not_available
vscode-tool-adapter adapter active not_available not_available unknown not yet measured not_available
ansa_builder agent active not_available scoped_repo_mutation unknown not yet measured codex exec --profile ansa-build --sandbox workspace-write
ansa_complex_builder agent active not_available scoped_repo_mutation unknown not yet measured codex exec --profile ansa-build --sandbox workspace-write` with a higher-reasoning build profile only when justified by shared-contract complexity
ansa_docs_researcher agent active not_available read_only unknown not yet measured codex exec --profile ansa-plan` or another read-only Codex research profile
ansa_explorer agent active not_available read_only unknown not yet measured codex exec --profile ansa-plan` or another read-only Codex planning or discovery profile
ansa_finaliser agent active not_available limited_github_mutation` only when PR/body/comment updates are explicitly authorised; otherwise `read_only unknown not yet measured codex exec --profile ansa-receipt` or a fresh Codex app session started with the receipt profile
ansa_hard_repairer agent active not_available scoped_repo_mutation unknown not yet measured codex exec --profile ansa-repair --sandbox workspace-write` with temporary stage escalation when needed
ansa_mechanic agent active not_available scoped_repo_mutation unknown not yet measured codex exec --profile ansa-build --sandbox workspace-write
ansa_planner agent active not_available read_only unknown not yet measured codex exec --profile ansa-plan` or a fresh Codex app session started with the planning profile
ansa_planner` plus `ansa_explorer` when target context must be mapped agent active not_available read_only unknown not yet measured codex exec --profile ansa-plan` or a fresh Codex app session started with the planning profile
ansa_repairer agent active not_available scoped_repo_mutation unknown not yet measured codex exec --profile ansa-repair --sandbox workspace-write
ansa_repairer agent active not_available scoped_repo_mutation` limited to conflict markers and required adjacent fixes unknown not yet measured codex exec --profile ansa-repair --sandbox workspace-write
ansa_reviewer agent active not_available read_only unknown not yet measured codex -p ansa-review review --uncommitted` or `codex review --model <operator-configured review model> --uncommitted
ansa_security_reviewer agent active not_available read_only unknown not yet measured codex -p ansa-review review --uncommitted` or `codex review --model <operator-configured review model> --uncommitted
ansa_verify_summariser agent active not_available read_only unknown not yet measured deterministic shell first; optional `codex exec --profile ansa-verify` only to summarise explicit failures after commands run
codex-skill-ansa generated_skill active not_available not_available unknown not yet measured $ansa
codex-skill-ansa-afk-goal-writer generated_skill active not_available not_available unknown not yet measured $ansa-afk-goal-writer
codex-skill-ansa-agent-feedback-review generated_skill active not_available not_available unknown not yet measured $ansa-agent-feedback-review
codex-skill-ansa-agent-run-launch-kit generated_skill active not_available not_available unknown not yet measured $ansa-agent-run-launch-kit
codex-skill-ansa-architecture-review generated_skill active not_available not_available unknown not yet measured $ansa-architecture-review
codex-skill-ansa-attack-review generated_skill active not_available not_available unknown not yet measured $ansa-attack-review
codex-skill-ansa-autonomous-pr-cycle generated_skill active not_available not_available unknown not yet measured $ansa-autonomous-pr-cycle
codex-skill-ansa-batch-run-planner generated_skill active not_available not_available unknown not yet measured $ansa-batch-run-planner
codex-skill-ansa-brain2-portfolio-memory-proposed-write generated_skill active not_available not_available unknown not yet measured $ansa-brain2-portfolio-memory-proposed-write
codex-skill-ansa-cartant-contact-status-sweep generated_skill active not_available not_available unknown not yet measured $ansa-cartant-contact-status-sweep
codex-skill-ansa-cartant-gmail-job-label-triage generated_skill active not_available not_available unknown not yet measured $ansa-cartant-gmail-job-label-triage
codex-skill-ansa-cartant-meeting-transcript-processing generated_skill active not_available not_available unknown not yet measured $ansa-cartant-meeting-transcript-processing
codex-skill-ansa-cartant-pre-meeting-brief generated_skill active not_available not_available unknown not yet measured $ansa-cartant-pre-meeting-brief
codex-skill-ansa-cartant-start-day generated_skill active not_available not_available unknown not yet measured $ansa-cartant-start-day
codex-skill-ansa-cartant-strategic-next-step-review generated_skill active not_available not_available unknown not yet measured $ansa-cartant-strategic-next-step-review
codex-skill-ansa-cartant-targeted-message-processing generated_skill active not_available not_available unknown not yet measured $ansa-cartant-targeted-message-processing
codex-skill-ansa-cartant-tracker-hygiene generated_skill active not_available not_available unknown not yet measured $ansa-cartant-tracker-hygiene
codex-skill-ansa-cartant-voice-note-processing generated_skill active not_available not_available unknown not yet measured $ansa-cartant-voice-note-processing
codex-skill-ansa-clean-baseline generated_skill active not_available not_available unknown not yet measured $ansa-clean-baseline
codex-skill-ansa-create-issue-from-evidence generated_skill active not_available not_available unknown not yet measured $ansa-create-issue-from-evidence
codex-skill-ansa-design-critique generated_skill active not_available not_available unknown not yet measured $ansa-design-critique
codex-skill-ansa-finalise-pr generated_skill active not_available not_available unknown not yet measured $ansa-finalise-pr
codex-skill-ansa-goose-controller-proof-prompt generated_skill active not_available not_available unknown not yet measured $ansa-goose-controller-proof-prompt
codex-skill-ansa-grill-me generated_skill active not_available not_available unknown not yet measured $ansa-grill-me
codex-skill-ansa-grill-with-docs generated_skill active not_available not_available unknown not yet measured $ansa-grill-with-docs
codex-skill-ansa-handoff generated_skill active not_available not_available unknown not yet measured $ansa-handoff
codex-skill-ansa-implement-issue generated_skill active not_available not_available unknown not yet measured $ansa-implement-issue
codex-skill-ansa-legal-document-review-response generated_skill active not_available not_available unknown not yet measured $ansa-legal-document-review-response
codex-skill-ansa-operator-prompt generated_skill active not_available not_available unknown not yet measured $ansa-operator-prompt
codex-skill-ansa-plan-issue generated_skill active not_available not_available unknown not yet measured $ansa-plan-issue
codex-skill-ansa-pr-approval-brief generated_skill active not_available not_available unknown not yet measured $ansa-pr-approval-brief
codex-skill-ansa-pr-review generated_skill active not_available not_available unknown not yet measured $ansa-pr-review
codex-skill-ansa-pr-tranche-closeout generated_skill active not_available not_available unknown not yet measured $ansa-pr-tranche-closeout
codex-skill-ansa-red-team generated_skill active not_available not_available unknown not yet measured $ansa-red-team
codex-skill-ansa-red-team-review-and-fix generated_skill active not_available not_available unknown not yet measured $ansa-red-team-review-and-fix
codex-skill-ansa-research generated_skill active not_available not_available unknown not yet measured $ansa-research
codex-skill-ansa-resolve-merge-rebase-conflicts generated_skill active not_available not_available unknown not yet measured $ansa-resolve-merge-rebase-conflicts
codex-skill-ansa-review-comment-closeout generated_skill active not_available not_available unknown not yet measured $ansa-review-comment-closeout
codex-skill-ansa-session-learning-review generated_skill active not_available not_available unknown not yet measured $ansa-session-learning-review
codex-skill-ansa-setup-pre-commit generated_skill active not_available not_available unknown not yet measured $ansa-setup-pre-commit
codex-skill-ansa-skillopt-skill-pilot generated_skill active not_available not_available unknown not yet measured $ansa-skillopt-skill-pilot
codex-skill-ansa-sync-local-repos-main generated_skill active not_available not_available unknown not yet measured $ansa-sync-local-repos-main
codex-skill-ansa-teach generated_skill active not_available not_available unknown not yet measured $ansa-teach
codex-skill-ansa-verify-pr generated_skill active not_available not_available unknown not yet measured $ansa-verify-pr
codex-skill-ansa-write-a-skill generated_skill active not_available not_available unknown not yet measured $ansa-write-a-skill
codex-skill-ansa-writing-for-agents generated_skill active not_available not_available unknown not yet measured $ansa-writing-for-agents
cartant-contact-status-sweep prompt active not_available not_available unknown not yet measured not_available
cartant-gmail-job-label-triage prompt active not_available not_available unknown not yet measured not_available
cartant-meeting-transcript-processing prompt active not_available not_available unknown not yet measured not_available
cartant-pre-meeting-brief prompt active not_available not_available unknown not yet measured not_available
cartant-start-day prompt active not_available not_available unknown not yet measured not_available
cartant-strategic-next-step-review prompt active not_available not_available unknown not yet measured not_available
cartant-targeted-message-processing prompt active not_available not_available unknown not yet measured not_available
cartant-tracker-hygiene prompt active not_available not_available unknown not yet measured not_available
cartant-voice-note-processing prompt active not_available not_available unknown not yet measured not_available
coding-afk-goal-writer prompt active not_available not_available unknown not yet measured not_available
coding-agent-run-launch-kit prompt active not_available not_available unknown not yet measured not_available
coding-ansa-operator-prompt prompt active not_available not_available unknown not yet measured not_available
coding-architecture-review prompt active not_available not_available unknown not yet measured not_available
coding-attack-review prompt active not_available not_available unknown not yet measured not_available
coding-autonomous-pr-cycle prompt active not_available not_available unknown not yet measured not_available
coding-batch-pr-supervisor-closeout prompt active not_available not_available unknown not yet measured not_available
coding-batch-run-planner prompt active not_available not_available unknown not yet measured not_available
coding-clean-baseline prompt active not_available not_available unknown not yet measured not_available
coding-corrective-pr-tranche prompt active not_available not_available unknown not yet measured not_available
coding-finalise-pr prompt active not_available not_available unknown not yet measured not_available
coding-goose-controller-proof-prompt prompt active not_available not_available unknown not yet measured not_available
coding-grill-with-docs prompt active not_available not_available unknown not yet measured not_available
coding-implement-issue prompt active not_available not_available unknown not yet measured not_available
coding-issue-plan prompt active not_available not_available unknown not yet measured not_available
coding-operator-decision-review-pack prompt active not_available not_available unknown not yet measured not_available
coding-pr-approval-brief prompt active not_available not_available unknown not yet measured not_available
coding-pr-body-repair prompt active not_available not_available unknown not yet measured not_available
coding-pr-review prompt active not_available not_available unknown not yet measured not_available
coding-pr-tranche-closeout prompt active not_available not_available unknown not yet measured not_available
coding-prototype prompt active not_available not_available unknown not yet measured not_available
coding-red-team-review-and-fix prompt active not_available not_available unknown not yet measured not_available
coding-research prompt active not_available not_available unknown not yet measured not_available
coding-resolve-merge-rebase-conflicts prompt active not_available not_available unknown not yet measured not_available
coding-review-comment-closeout prompt active not_available not_available unknown not yet measured not_available
coding-setup-pre-commit prompt active not_available not_available unknown not yet measured not_available
coding-stale-pr-sweeper-report prompt active not_available not_available unknown not yet measured not_available
coding-sync-local-repos-main prompt active not_available not_available unknown not yet measured not_available
coding-to-spec prompt active not_available not_available unknown not yet measured not_available
coding-to-tickets prompt active not_available not_available unknown not yet measured not_available
coding-triage prompt active not_available not_available unknown not yet measured not_available
coding-verify-pr prompt active not_available not_available unknown not yet measured not_available
github-pr-review-brain2-chatgpt prompt active not_available not_available unknown not yet measured not_available
github-pr-review-chatgpt prompt active not_available not_available unknown not yet measured not_available
opencode-stage-builder prompt active not_available not_available unknown not yet measured not_available
opencode-stage-fixer prompt active not_available not_available unknown not yet measured not_available
opencode-stage-planner prompt active not_available not_available unknown not yet measured not_available
opencode-stage-receipt-finaliser prompt active not_available not_available unknown not yet measured not_available
opencode-stage-reviewer prompt active not_available not_available unknown not yet measured not_available
opencode-stage-verifier-reviewer prompt active not_available not_available unknown not yet measured not_available
operations-agent-feedback-review prompt active not_available not_available unknown not yet measured not_available
operations-brain2-portfolio-memory-proposed-write prompt active not_available not_available unknown not yet measured not_available
operations-create-issue-from-evidence prompt active not_available not_available unknown not yet measured not_available
operations-session-learning-review prompt active not_available not_available unknown not yet measured not_available
operations-skillopt-skill-pilot prompt active not_available not_available unknown not yet measured not_available
productivity-design-critique prompt active not_available not_available unknown not yet measured not_available
productivity-gmail-inbox-management prompt active not_available not_available unknown not yet measured not_available
productivity-grill-me prompt active not_available not_available unknown not yet measured not_available
productivity-handoff prompt active not_available not_available unknown not yet measured not_available
productivity-legal-document-review-response prompt active not_available not_available unknown not yet measured not_available
productivity-red-team prompt active not_available not_available unknown not yet measured not_available
productivity-teach prompt active not_available not_available unknown not yet measured not_available
productivity-wayfinder prompt active not_available not_available unknown not yet measured not_available
productivity-write-a-skill prompt active not_available not_available unknown not yet measured not_available
productivity-writing-for-agents prompt active not_available not_available unknown not yet measured not_available
session-closeout-brain2-chatgpt prompt active not_available not_available unknown not yet measured not_available
business-development.opportunity-briefing route active business-development read_only unknown not yet measured not_available
cartant.contact-status-sweep route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.contact-status-sweep
cartant.gmail-job-label-triage route active cartant read_only unknown not yet measured Use Ansa. Route: cartant.gmail-job-label-triage
cartant.meeting-transcript-processing route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.meeting-transcript-processing
cartant.pre-meeting-brief route active cartant read_only unknown not yet measured Use Ansa. Route: cartant.pre-meeting-brief
cartant.start-day route active cartant read_only unknown not yet measured Use Ansa. Route: cartant.start-day
cartant.strategic-next-step-review route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.strategic-next-step-review
cartant.targeted-message-processing route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.targeted-message-processing
cartant.tracker-hygiene route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.tracker-hygiene
cartant.voice-note-processing route active cartant mixed unknown not yet measured Use Ansa. Route: cartant.voice-note-processing
coding.afk-goal-writer route active coding read_only unknown not yet measured Use Ansa. Route: coding.afk-goal-writer
coding.agent-run-launch-kit route active coding read_only unknown not yet measured Use Ansa. Route: coding.agent-run-launch-kit
coding.ansa-operator-prompt route active coding read_only field_tested not yet measured Use Ansa. Route: coding.ansa-operator-prompt
coding.architecture-review route active coding mixed unknown not yet measured Use Ansa. Route: coding.architecture-review
coding.attack-review route active coding read_only field_tested not yet measured Use Ansa. Route: coding.attack-review
coding.autonomous-pr-cycle route active coding mixed field_tested not yet measured Use Ansa. Route: coding.autonomous-pr-cycle
coding.batch-run-planner route active coding read_only unknown not yet measured Use Ansa. Route: coding.batch-run-planner
coding.clean-baseline route active coding mixed unknown not yet measured Use Ansa. Route: coding.clean-baseline
coding.finalise-pr route active coding read_only field_tested not yet measured Use Ansa. Route: coding.finalise-pr
coding.goose-controller-proof-prompt route active coding read_only unknown not yet measured Use Ansa. Route: coding.goose-controller-proof-prompt
coding.grill-with-docs route active coding mixed unknown not yet measured Use Ansa. Route: coding.grill-with-docs
coding.implement-issue route active coding mutating field_tested not yet measured Use Ansa. Route: coding.implement-issue
coding.plan-issue route active coding read_only unknown not yet measured Use Ansa. Route: coding.plan-issue
coding.pr-review route active coding read_only unknown not yet measured Use Ansa. Route: coding.pr-review
coding.prototype route active coding mixed unknown not yet measured Use Ansa. Route: coding.prototype
coding.red-team-fix-prompt route active coding read_only unknown not yet measured Use Ansa. Route: coding.red-team-fix-prompt
coding.repo-hygiene-cleanup route active coding mixed unknown not yet measured not_available
coding.research route active coding mixed unknown not yet measured Use Ansa. Route: coding.research
coding.resolve-merge-rebase-conflicts route active coding mixed unknown not yet measured Use Ansa. Route: coding.resolve-merge-rebase-conflicts
coding.review-comment-closeout route active coding mixed field_tested not yet measured Use Ansa. Route: coding.review-comment-closeout
coding.safe-branch-main-reset route active coding mixed unknown not yet measured not_available
coding.setup-pre-commit route active coding mixed unknown not yet measured Use Ansa. Route: coding.setup-pre-commit
coding.sync-local-repos-main route active coding mixed unknown not yet measured Use Ansa. Route: coding.sync-local-repos-main
coding.to-spec route active coding mixed unknown not yet measured Use Ansa. Route: coding.to-spec
coding.to-tickets route active coding mixed unknown not yet measured Use Ansa. Route: coding.to-tickets
coding.triage route active coding read_only unknown not yet measured Use Ansa. Route: coding.triage
coding.unattended-build-loop route active coding mixed unknown not yet measured not_available
coding.verify-pr route active coding read_only field_tested not yet measured Use Ansa. Route: coding.verify-pr
household-shopping.child-clothing-shopping-plan route active household-shopping read_only unknown not yet measured not_available
household-shopping.shopping-agent-plan route active household-shopping read_only unknown not yet measured not_available
operations.agent-feedback-review route active operations mixed unknown not yet measured Use Ansa. Route: operations.agent-feedback-review
operations.ansa-self-maintenance route active operations mixed unknown not yet measured not_available
operations.ansa-self-review route active operations read_only unknown not yet measured not_available
operations.brain2-portfolio-memory-proposed-write route active operations mixed unknown not yet measured Use Ansa. Route: operations.brain2-portfolio-memory-proposed-write
operations.closed-issue-audit route active operations read_only unknown not yet measured not_available
operations.create-issue-from-evidence route active operations mixed pilot not yet measured Use Ansa. Route: operations.create-issue-from-evidence
operations.cross-repo-receipt-review route active operations read_only unknown not yet measured not_available
operations.outcome-learning-review route active operations read_only unknown not yet measured not_available
operations.post-merge-bug-sweep route active operations read_only pilot not yet measured not_available
operations.receipt-learning-review route active operations read_only unknown not yet measured not_available
operations.session-learning-review route active operations read_only unknown not yet measured Use Ansa. Route: operations.session-learning-review
operations.skillopt-skill-pilot route active operations mixed pilot not yet measured Use Ansa. Route: operations.skillopt-skill-pilot
planning.event-planning route active planning read_only unknown not yet measured not_available
planning.trip-planning route active planning read_only unknown not yet measured not_available
productivity.design-critique route active productivity read_only pilot not yet measured Use Ansa. Route: productivity.design-critique
productivity.gmail-inbox-management route active productivity mixed draft not yet measured Use Ansa. Route: productivity.gmail-inbox-management
productivity.grill-me route active productivity read_only unknown not yet measured Use Ansa. Route: productivity.grill-me
productivity.handoff route active productivity mixed unknown not yet measured Use Ansa. Route: productivity.handoff
productivity.legal-document-review-response route active productivity mixed draft not yet measured Use Ansa. Route: productivity.legal-document-review-response
productivity.red-team route active productivity read_only unknown not yet measured Use Ansa. Route: productivity.red-team
productivity.teach route active productivity mixed unknown not yet measured Use Ansa. Route: productivity.teach
productivity.to-questionnaire route active productivity mixed unknown not yet measured not_available
productivity.wait-what route active productivity read_only unknown not yet measured not_available
productivity.wayfinder route active productivity mixed pilot not yet measured Use Ansa. Route: productivity.wayfinder
productivity.write-a-skill route active productivity mixed unknown not yet measured Use Ansa. Route: productivity.write-a-skill
productivity.writing-for-agents route active productivity mixed unknown not yet measured Use Ansa. Route: productivity.writing-for-agents
transcription.meeting-transcript-decomposition route active transcription read_only unknown not yet measured not_available
transcription.mono-mic-session-transcript route active transcription read_only unknown not yet measured not_available
transcription.pixel-recording-transcript route active transcription read_only unknown not yet measured not_available
assert-pr-review-threads-closed script active not_available read_only unknown not yet measured not_available
check-enki-entrypoint-card script active not_available read_only unknown not yet measured not_available
check-routing-conformance script active not_available read_only unknown not yet measured not_available
collect-outcome-evidence script active not_available mixed unknown not yet measured not_available
export-agent-pack script active not_available mixed unknown not yet measured not_available
generate-active-catalogue script active not_available read_only unknown not yet measured not_available
generate-ansa-scorecard script active not_available mixed unknown not yet measured not_available
generate-codex-skill-wrappers script active not_available mixed unknown not yet measured not_available
generate-codex-token-waste-report script active not_available mixed unknown not yet measured not_available
generate-evidence-console script active not_available mixed unknown not yet measured not_available
generate-evidence-console-dashboard script active not_available mixed unknown not yet measured not_available
generate-outcome-learning-report script active not_available mixed unknown not yet measured not_available
generate-self-improvement-candidates script active not_available mixed unknown not yet measured not_available
install-codex-skill-sync-hook script active not_available mixed unknown not yet measured not_available
open-review-pr script active not_available mixed unknown not yet measured not_available
opencode-preflight script active not_available mixed unknown not yet measured not_available
prepare-review-thread-graphql-vars script active not_available read_only unknown not yet measured not_available
prepare-self-improvement-pr-evidence script active not_available mixed unknown not yet measured not_available
produce-codex-token-waste-summaries script active not_available mixed unknown not yet measured not_available
project-active-catalogue-to-notion script active not_available mixed unknown not yet measured not_available
read-pr-status script active not_available read_only unknown not yet measured not_available
render-operator-prompt script active not_available read_only unknown not yet measured not_available
run-conformance-package script active not_available read_only unknown not yet measured not_available
run-eval script active not_available mixed unknown not yet measured not_available
run-self-improvement-loop script active not_available mixed unknown not yet measured not_available
run-weekly-codex-token-waste-report script active not_available mixed unknown not yet measured not_available
safe-update-pr-body script active not_available mixed unknown not yet measured not_available
sync-local-repos-main script active not_available mixed unknown not yet measured not_available
sync-mattpocock script active not_available mixed unknown not yet measured not_available
update-sticky-pr-comment script active not_available mixed unknown not yet measured not_available
validate-library script active not_available read_only unknown not yet measured not_available
validate-pr-body script active not_available read_only unknown not yet measured not_available
validate-pr-formation-decision script active not_available read_only unknown not yet measured not_available
validate-validation-evidence script active not_available read_only unknown not yet measured not_available
evidence-first-intake skill active not_available not_available pilot not yet measured not_available
github-pr-review skill active not_available not_available field_tested not yet measured not_available
grilling skill active not_available not_available pilot not yet measured not_available
metadata-integrity skill active not_available not_available pilot not yet measured not_available
proposed-write-gate skill active not_available not_available pilot not yet measured not_available
record-deduplication-and-linking skill active not_available not_available pilot not yet measured not_available
repo-aware-debug-repair skill active not_available not_available pilot not yet measured not_available
review-participant-gate skill active not_available not_available pilot not yet measured not_available
session-closeout-governance skill active not_available not_available pilot not yet measured not_available
skill-intake-normalise-register skill active not_available not_available pilot not yet measured not_available
source-of-truth-resolution skill active not_available not_available pilot not yet measured not_available
test-driven-execution skill active not_available not_available pilot not yet measured not_available
adversarial-validator-matrix template active not_available not_available unknown not yet measured not_available
afk-pr-settlement-contract template active not_available not_available unknown not yet measured not_available
agent-feedback-event template active not_available not_available unknown not yet measured not_available
ansa-conformance-workflow template active not_available not_available unknown not yet measured not_available
ansa-sticky-comment template active not_available not_available unknown not yet measured not_available
approved-tranche-execution template active not_available not_available unknown not yet measured not_available
assertion-ledger template active not_available not_available unknown not yet measured not_available
audit-marker-contract template active not_available not_available unknown not yet measured not_available
batch-closeout-report template active not_available not_available unknown not yet measured not_available
brain2-context-use-check template active not_available not_available unknown not yet measured not_available
brain2-portfolio-memory-proposal template active not_available not_available unknown not yet measured not_available
closed-issue-audit-report template active not_available not_available unknown not yet measured not_available
closeout-gap-check template active not_available not_available unknown not yet measured not_available
coding-assurance-gate template active not_available not_available unknown not yet measured not_available
compact-stage-handoff template active not_available not_available unknown not yet measured not_available
component-impact-report template active not_available not_available unknown not yet measured not_available
context-pack template active not_available not_available unknown not yet measured not_available
decision-action-extract template active not_available not_available unknown not yet measured not_available
destructive-action-approval template active not_available not_available unknown not yet measured not_available
enki-entrypoint-card template active not_available not_available unknown not yet measured not_available
enki-entrypoint-preflight template active not_available not_available unknown not yet measured not_available
enki-handoff template active not_available not_available unknown not yet measured not_available
execution-contract template active not_available not_available unknown not yet measured not_available
final-summary template active not_available not_available unknown not yet measured not_available
finding-lifecycle-contract template active not_available not_available unknown not yet measured not_available
generated-component-proposal template active not_available not_available unknown not yet measured not_available
generated-output-refresh template active not_available not_available unknown not yet measured not_available
git-cleanup-dry-run-report template active not_available not_available unknown not yet measured not_available
github-review-surface-readback template active not_available not_available unknown not yet measured not_available
maintenance-report template active not_available not_available unknown not yet measured not_available
manual-review-report template active not_available not_available unknown not yet measured not_available
meeting-decomposition template active not_available not_available unknown not yet measured not_available
multi-stage-run-receipt template active not_available not_available unknown not yet measured not_available
operator-constitution-spec template active not_available not_available unknown not yet measured not_available
operator-decision-review-pack template active not_available not_available unknown not yet measured not_available
opportunity-brief template active not_available not_available unknown not yet measured not_available
outcome-learning-report template active not_available not_available unknown not yet measured not_available
owned-skill-checklist template active not_available not_available unknown not yet measured not_available
planning-evidence-table template active not_available not_available unknown not yet measured not_available
planning-intake template active not_available not_available unknown not yet measured not_available
portfolio-adoption-status-report template active not_available not_available unknown not yet measured not_available
portfolio-repo-registry template active not_available not_available unknown not yet measured not_available
post-merge-bug-sweep-report template active not_available not_available unknown not yet measured not_available
pr-ansa-usage template active not_available not_available unknown not yet measured not_available
pr-formation-decision template active not_available not_available unknown not yet measured not_available
pr-handoff-readiness template active not_available not_available unknown not yet measured not_available
proposed-writes template active not_available not_available unknown not yet measured not_available
provider-model-registry template active not_available not_available unknown not yet measured not_available
raw-source-metadata-table template active not_available not_available unknown not yet measured not_available
receipt-learning-report template active not_available not_available unknown not yet measured not_available
red-team-report template active not_available not_available unknown not yet measured not_available
repo-starter-ci-governance-notes template active not_available not_available unknown not yet measured not_available
repo-starter-clean-room-external-inspiration template active not_available not_available unknown not yet measured not_available
repo-starter-coding-standards template active not_available not_available unknown not yet measured not_available
repo-starter-consumer-repo-pointer template active not_available not_available unknown not yet measured not_available
repo-starter-governance-conformance-checklist template active not_available not_available unknown not yet measured not_available
repo-starter-governance-repo-adapter template active not_available not_available unknown not yet measured not_available
repo-starter-issue-template-bug template active not_available not_available unknown not yet measured not_available
repo-starter-issue-template-docs template active not_available not_available unknown not yet measured not_available
repo-starter-issue-template-feature template active not_available not_available unknown not yet measured not_available
repo-starter-pr-template template active not_available not_available unknown not yet measured not_available
repo-starter-readme template active not_available not_available unknown not yet measured not_available
repo-starter-red-team-checklist template active not_available not_available unknown not yet measured not_available
repo-starter-repo-adapter-notes template active not_available not_available unknown not yet measured not_available
repo-starter-repo-constitution template active not_available not_available unknown not yet measured not_available
repo-starter-review-checklist template active not_available not_available unknown not yet measured not_available
repo-starter-validation-evidence template active not_available not_available unknown not yet measured not_available
review-participant-gate template active not_available not_available unknown not yet measured not_available
review-summary template active not_available not_available unknown not yet measured not_available
run-receipt template active not_available not_available unknown not yet measured not_available
runtime-proof template active not_available not_available unknown not yet measured not_available
self-review-report template active not_available not_available unknown not yet measured not_available
skill-contract template active not_available not_available unknown not yet measured not_available
skill-intake-report template active not_available not_available unknown not yet measured not_available
stage-contract template active not_available not_available unknown not yet measured not_available
stale-pr-sweeper-report template active not_available not_available unknown not yet measured not_available
stateful-skill-action-profile template active not_available not_available unknown not yet measured not_available
stateful-skill-approval-lifecycle template active not_available not_available unknown not yet measured not_available
stateful-skill-brain2-adapter-contract template active not_available not_available unknown not yet measured not_available
stateful-skill-state-proposal template active not_available not_available unknown not yet measured not_available
stop-report template active not_available not_available unknown not yet measured not_available
task-contract template active not_available not_available unknown not yet measured not_available
tool-bakeoff-run template active not_available not_available unknown not yet measured not_available
tool-bakeoff-summary template active not_available not_available unknown not yet measured not_available
tranche-manifest template active not_available not_available unknown not yet measured not_available
transcript-decomposition template active not_available not_available unknown not yet measured not_available
unattended-build-loop-manifest template active not_available not_available unknown not yet measured not_available
validation-evidence template active not_available not_available unknown not yet measured not_available
wayfinder-mission-run-brief template active not_available not_available unknown not yet measured not_available
write-receipt template active not_available not_available unknown not yet measured not_available
debug-repair-validate workflow active not_available not_available unknown not yet measured not_available
review-open-prs workflow active not_available not_available unknown not yet measured not_available
skill-intake workflow active not_available not_available unknown not yet measured not_available

Dashboard Readiness

KPI Evidence Status Measured Criteria
catalogue_coverage fact partially_supported not yet measured unknown
projection_reliability not yet measured unknown not yet measured unknown
receipt_coverage derived unmet_or_insufficient 56.6% outcome-evidence coverage in selected scorecard window unknown
usage_attribution_quality derived partially_supported observed_usage_without_complete_exact_totals unknown
scorecard_sample_size fact met 112 unknown
learning_loop_maturity fact supported_with_candidates active_candidates=9; resolved_candidates=2 unknown
actionability_test derived partially_supported repo_native_snapshot unknown
low_maintenance_test fact met repo_native_generation unknown

Improvement Queue

Active candidates: 9 | Resolved candidates: 2 | Draft PR proposals: 0 | Operator decision required: yes

Measured PRs and receipts above describe the full evidence base, not just unresolved candidate items.

  • missed_component-0aab406a9428: update_workflow_template_adapter_guidance for realms/coding/workflows/attack-review/SKILL.md high blocked
  • missing_runtime_proof-a90d20266ff0: update_workflow_template_adapter_guidance for unknown high blocked
  • missing_validation-1363fe34e513: strengthen_validator for unknown high blocked
  • review_burden-16a7eaabcd1b: update_workflow_template_adapter_guidance for unknown high blocked
  • rework-3fb31b1061ab: update_workflow_template_adapter_guidance for library/templates/portfolio-repo-registry.yml high blocked
  • scope_creep-d64f4b08cb35: update_workflow_template_adapter_guidance for realms/coding/workflows/implement-issue/SKILL.md high blocked
  • sequencing_problem-b762366e22bb: amend_existing_issue for coding.autonomous-pr-cycle high blocked
  • stale_canon-4859578a9a3e: update_workflow_template_adapter_guidance for realms/coding/workflows/finalise-pr/SKILL.md high blocked
  • weak_scope-7b6b235e81d7: update_workflow_template_adapter_guidance for unknown high blocked

Candidate Decision Queue

missed_component-0aab406a9428 missed_component realms/coding/workflows/attack-review/SKILL.md 19 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Failure class is not allowlisted for automation draft PR proposals. approve, defer, waive
missing_runtime_proof-a90d20266ff0 missing_runtime_proof unknown 4 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Failure class is not allowlisted for automation draft PR proposals. approve, defer, waive
missing_validation-1363fe34e513 missing_validation unknown 5 high high strengthen_validator blocked Strengthen deterministic validator for repeated self-improvement misses none Automation draft PR proposals require a concrete target surface and candidate target component evidence. approve, defer, waive
review_burden-16a7eaabcd1b review_burden unknown 4 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Automation draft PR proposals require a concrete target surface and candidate target component evidence. approve, defer, waive
rework-3fb31b1061ab rework library/templates/portfolio-repo-registry.yml 5 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Automation draft PR proposals require candidate target component evidence (`ansa_components.candidate_target_components`). approve, defer, waive
scope_creep-d64f4b08cb35 scope_creep realms/coding/workflows/implement-issue/SKILL.md 37 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Automation draft PR proposals require candidate target component evidence (`ansa_components.candidate_target_components`). approve, defer, waive
sequencing_problem-b762366e22bb sequencing_problem coding.autonomous-pr-cycle 10 high high amend_existing_issue blocked Amend existing corrective issue with repeated self-improvement evidence none Recommended action amend_existing_issue requires human review instead of automation draft PR generation. approve, defer, waive
stale_canon-4859578a9a3e stale_canon realms/coding/workflows/finalise-pr/SKILL.md 12 high high update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Failure class is not allowlisted for automation draft PR proposals. approve, defer, waive
weak_scope-7b6b235e81d7 weak_scope unknown 6 high medium update_workflow_template_adapter_guidance blocked Tighten workflow, template, or adapter guidance for repeated self-improvement misses none Automation draft PR proposals require a concrete target surface and candidate target component evidence. approve, defer, waive

Copy-Paste Action Packs

missed_component-0aab406a9428 blocked

Approve the issue proposal if you want a bounded follow-up recorded for `realms/coding/workflows/attack-review/SKILL.md`; otherwise defer or waive.

  • Problem: Repeated evidence says missed component is still causing follow-through friction around `realms/coding/workflows/attack-review/SKILL.md`.
  • Follow-up: Record a bounded GitHub issue for `realms/coding/workflows/attack-review/SKILL.md` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: missed_component-0aab406a9428
failure_class: missed_component
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: realms/coding/workflows/attack-review/SKILL.md
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60missed_component%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60missed_component-0aab406a9428%60%0A-%20Evidence%20count%3A%2019%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20639%2C%20534%2C%20483%2C%20357%2C%20325%2C%20321%2C%20260%2C%20259%2C%20257%2C%20195%2C%20178%2C%20146%2C%20139%2C%20138%2C%20142%0A-%20Affected%20receipts%3A%20%60runs%2F2026-06-25-issue-633-session-learning-review.yml%60%2C%20%60runs%2F2026-06-24%2Fissue-607-enki-entrypoint-evidence%2Frun-receipt.yml%60%2C%20%60runs%2F2026-06-20-issue-531-pr-tranche-closeout.yml%60%2C%20%60runs%2F2026-06-19-issue-221-pr-review-evidence-discovery%2Frun-receipt.yml%60%2C%20%60runs%2F2026-06-14-issue-401-planning-depth-modes.yml%60%2C%20%60runs%2F2026-06-08-issues-339-341-332-workflow-control.yml%60%2C%20%60runs%2F2026-06-04-pr-325-planning-workflows.yml%60%2C%20%60runs%2F2026-06-03-issue-313-downstream-validator-checkout-token.yml%60%2C%20%60runs%2F2026-05-24-issue-255-finding-lifecycle-contract.yml%60%2C%20%60runs%2F2026-05-24-issue-254-audit-marker-contract.yml%60%2C%20%60runs%2F2026-05-24-issue-253-portfolio-registry-blocked.yml%60%2C%20%60runs%2F2026-05-19-issue-191-agent-feedback-event.yml%60%2C%20%60runs%2F2026-05-18-issue-135-capped-api-overflow-gateway.yml%60%2C%20%60runs%2F2026-05-17-issue-93-transcription-session-workflows.yml%60%2C%20%60runs%2F2026-05-17-issue-134-stage-model-routing.yml%60%2C%20%60runs%2F2026-05-17-issue-133-batch-closeout-report.yml%60%2C%20%60runs%2F2026-05-17-issue-131-adapter-cost-policy.yml%60%2C%20%60runs%2F2026-05-14-issue-74-cartant-adapters-reference.yml%60%2C%20%60runs%2F2026-05-14-ansa-repo-audit-report.yml%60%0A-%20Affected%20components%3A%20%60realms%2Fcoding%2Fworkflows%2Fattack-review%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fimplement-issue%2FSKILL.md%60%2C%20%60coding.autonomous-pr-cycle%60%2C%20%60coding.implement-issue%60%2C%20%60library%2Fadapters%2Fchatgpt%2Fusage.md%60%2C%20%60library%2Fadapters%2Fcost-policy.md%60%2C%20%60library%2Fadapters%2Flitellm-openrouter-gateway.md%60%2C%20%60library%2Fadapters%2Fmodel-routing-standard.md%60%2C%20%60library%2Fprompts%2Fchatgpt%2Fgithub-pr-review.md%60%2C%20%60library%2Ftemplates%2Faudit-marker-contract.yml%60%2C%20%60library%2Ftemplates%2Fbatch-closeout-report.md%60%2C%20%60library%2Ftemplates%2Fcompact-stage-handoff.md%60%2C%20%60library%2Ftemplates%2Ffinding-lifecycle-contract.yml%60%2C%20%60library%2Ftemplates%2Fplanning-evidence-table.md%60%2C%20%60library%2Ftemplates%2Fportfolio-repo-registry.yml%60%2C%20%60library%2Ftemplates%2Ftranscript-decomposition.md%60%2C%20%60realms%2Fcoding%2Frun-desks%2Fpr-tranche-closeout%2FREADME.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fautonomous-pr-cycle%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Ffinalise-pr%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fpr-review%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fverify-pr%2FSKILL.md%60%2C%20%60realms%2Foperations%2Fworkflows%2Fagent-feedback-review%2FSKILL.md%60%2C%20%60realms%2Foperations%2Fworkflows%2Fsession-learning-review%2FSKILL.md%60%0A-%20New%20component%20needed%3A%20true%0A-%20Proposed%20target%20surface%3A%20%60realms%2Fcoding%2Fworkflows%2Fattack-review%2FSKILL.md%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%2C%20%60ansa_components.workflows_used%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%2019%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60missed_component%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - missed_component-0aab406a9428
  failure_classes:
    - missed_component
  target_surfaces:
    - realms/coding/workflows/attack-review/SKILL.md
  covered_receipts:
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-19-issue-221-pr-review-evidence-discovery/run-receipt.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-04-pr-325-planning-workflows.yml
    - runs/2026-06-03-issue-313-downstream-validator-checkout-token.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-24-issue-254-audit-marker-contract.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-18-issue-135-capped-api-overflow-gateway.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
    - runs/2026-05-17-issue-131-adapter-cost-policy.yml
    - runs/2026-05-14-issue-74-cartant-adapters-reference.yml
    - runs/2026-05-14-ansa-repo-audit-report.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Failure class is not allowlisted for automation draft PR proposals.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: missed_component-0aab406a9428
failure_class: missed_component
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missed_component-0aab406a9428
  failure_classes:
    - missed_component
  target_surfaces:
    - realms/coding/workflows/attack-review/SKILL.md
  covered_receipts:
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-19-issue-221-pr-review-evidence-discovery/run-receipt.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-04-pr-325-planning-workflows.yml
    - runs/2026-06-03-issue-313-downstream-validator-checkout-token.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-24-issue-254-audit-marker-contract.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-18-issue-135-capped-api-overflow-gateway.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
    - runs/2026-05-17-issue-131-adapter-cost-policy.yml
    - runs/2026-05-14-issue-74-cartant-adapters-reference.yml
    - runs/2026-05-14-ansa-repo-audit-report.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: missed_component-0aab406a9428
failure_class: missed_component
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missed_component-0aab406a9428
  failure_classes:
    - missed_component
  target_surfaces:
    - realms/coding/workflows/attack-review/SKILL.md
  covered_receipts:
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-19-issue-221-pr-review-evidence-discovery/run-receipt.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-04-pr-325-planning-workflows.yml
    - runs/2026-06-03-issue-313-downstream-validator-checkout-token.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-24-issue-254-audit-marker-contract.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-18-issue-135-capped-api-overflow-gateway.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
    - runs/2026-05-17-issue-131-adapter-cost-policy.yml
    - runs/2026-05-14-issue-74-cartant-adapters-reference.yml
    - runs/2026-05-14-ansa-repo-audit-report.yml
  reason: <fill in operator-approved reason>
missing_runtime_proof-a90d20266ff0 blocked

Approve the issue proposal if you want a bounded follow-up recorded for `unknown`; otherwise defer or waive.

  • Problem: Repeated evidence says missing runtime proof is still causing follow-through friction around `unknown`.
  • Follow-up: Record a bounded GitHub issue for `unknown` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: missing_runtime_proof-a90d20266ff0
failure_class: missing_runtime_proof
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: unknown
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60missing_runtime_proof%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60missing_runtime_proof-a90d20266ff0%60%0A-%20Evidence%20count%3A%204%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20348%2C%20347%2C%20166%2C%20116%0A-%20Affected%20receipts%3A%20%60runs%2F2026-06-07-issue-337-pr-body-contract.yml%60%2C%20%60runs%2F2026-06-07-issue-344-context-loading-budget.yml%60%2C%20%60runs%2F2026-05-18-issue-164-165-self-improvement-loop-automation.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%0A-%20Affected%20components%3A%20unknown%0A-%20New%20component%20needed%3A%20false%0A-%20Proposed%20target%20surface%3A%20%60unknown%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%204%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60missing_runtime_proof%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - missing_runtime_proof-a90d20266ff0
  failure_classes:
    - missing_runtime_proof
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-06-07-issue-344-context-loading-budget.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Failure class is not allowlisted for automation draft PR proposals.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: missing_runtime_proof-a90d20266ff0
failure_class: missing_runtime_proof
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missing_runtime_proof-a90d20266ff0
  failure_classes:
    - missing_runtime_proof
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-06-07-issue-344-context-loading-budget.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: missing_runtime_proof-a90d20266ff0
failure_class: missing_runtime_proof
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missing_runtime_proof-a90d20266ff0
  failure_classes:
    - missing_runtime_proof
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-06-07-issue-344-context-loading-budget.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
  reason: <fill in operator-approved reason>
missing_validation-1363fe34e513 blocked

Approve the issue proposal if you want a bounded follow-up recorded for `unknown`; otherwise defer or waive.

  • Problem: Repeated evidence says missing validation is still causing follow-through friction around `unknown`.
  • Follow-up: Record a bounded GitHub issue for `unknown` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: missing_validation-1363fe34e513
failure_class: missing_validation
issue_title: Strengthen deterministic validator for repeated self-improvement misses
target_surface: unknown
github_issue_url: https://github.com/bigfeet-boy/ansa/issues/new?title=Strengthen%20deterministic%20validator%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60missing_validation%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60missing_validation-1363fe34e513%60%0A-%20Evidence%20count%3A%205%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20187%2C%20166%2C%20116%2C%20106%2C%20104%0A-%20Affected%20receipts%3A%20%60runs%2F2026-05-19-issue-10-minimal-agent-pack-export.yml%60%2C%20%60runs%2F2026-05-18-issue-164-165-self-improvement-loop-automation.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%2C%20%60runs%2F2026-05-14-issue-78-impact-scorecard.yml%60%2C%20%60runs%2F2026-05-14-issue-77-outcome-evidence.yml%60%0A-%20Affected%20components%3A%20unknown%0A-%20New%20component%20needed%3A%20false%0A-%20Proposed%20target%20surface%3A%20%60unknown%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%205%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60missing_validation%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Add%20or%20tighten%20a%20deterministic%20validator%20that%20catches%20the%20repeated%20failure%20pattern%20with%20low%20false-positive%20risk.%0A-%20Cover%20the%20validator%20with%20tests%20proving%20weak%20evidence%20does%20not%20trigger%20enforcement.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - missing_validation-1363fe34e513
  failure_classes:
    - missing_validation
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Automation draft PR proposals require a concrete target surface and candidate target component evidence.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: missing_validation-1363fe34e513
failure_class: missing_validation
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missing_validation-1363fe34e513
  failure_classes:
    - missing_validation
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: missing_validation-1363fe34e513
failure_class: missing_validation
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - missing_validation-1363fe34e513
  failure_classes:
    - missing_validation
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
review_burden-16a7eaabcd1b blocked

Approve the issue proposal if you want a bounded follow-up recorded for `unknown`; otherwise defer or waive.

  • Problem: Repeated evidence says review burden is still causing follow-through friction around `unknown`.
  • Follow-up: Record a bounded GitHub issue for `unknown` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: review_burden-16a7eaabcd1b
failure_class: review_burden
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: unknown
github_issue_url: https://github.com/bigfeet-boy/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60review_burden%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60review_burden-16a7eaabcd1b%60%0A-%20Evidence%20count%3A%204%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20187%2C%20166%2C%20116%2C%20104%0A-%20Affected%20receipts%3A%20%60runs%2F2026-05-19-issue-10-minimal-agent-pack-export.yml%60%2C%20%60runs%2F2026-05-18-issue-164-165-self-improvement-loop-automation.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%2C%20%60runs%2F2026-05-14-issue-77-outcome-evidence.yml%60%0A-%20Affected%20components%3A%20unknown%0A-%20New%20component%20needed%3A%20false%0A-%20Proposed%20target%20surface%3A%20%60unknown%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%204%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60review_burden%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - review_burden-16a7eaabcd1b
  failure_classes:
    - review_burden
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Automation draft PR proposals require a concrete target surface and candidate target component evidence.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: review_burden-16a7eaabcd1b
failure_class: review_burden
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - review_burden-16a7eaabcd1b
  failure_classes:
    - review_burden
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: review_burden-16a7eaabcd1b
failure_class: review_burden
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - review_burden-16a7eaabcd1b
  failure_classes:
    - review_burden
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
rework-3fb31b1061ab blocked

Approve the issue proposal if you want a bounded follow-up recorded for `library/templates/portfolio-repo-registry.yml`; otherwise defer or waive.

  • Problem: Repeated evidence says rework is still causing follow-through friction around `library/templates/portfolio-repo-registry.yml`.
  • Follow-up: Record a bounded GitHub issue for `library/templates/portfolio-repo-registry.yml` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: rework-3fb31b1061ab
failure_class: rework
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: library/templates/portfolio-repo-registry.yml
github_issue_url: https://github.com/bigfeet-boy/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60rework%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60rework-3fb31b1061ab%60%0A-%20Evidence%20count%3A%205%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20257%2C%20187%2C%20166%2C%20116%2C%20104%0A-%20Affected%20receipts%3A%20%60runs%2F2026-05-24-issue-253-portfolio-registry-blocked.yml%60%2C%20%60runs%2F2026-05-19-issue-10-minimal-agent-pack-export.yml%60%2C%20%60runs%2F2026-05-18-issue-164-165-self-improvement-loop-automation.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%2C%20%60runs%2F2026-05-14-issue-77-outcome-evidence.yml%60%0A-%20Affected%20components%3A%20%60library%2Ftemplates%2Fportfolio-repo-registry.yml%60%0A-%20New%20component%20needed%3A%20true%0A-%20Proposed%20target%20surface%3A%20%60library%2Ftemplates%2Fportfolio-repo-registry.yml%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%205%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60rework%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - rework-3fb31b1061ab
  failure_classes:
    - rework
  target_surfaces:
    - library/templates/portfolio-repo-registry.yml
  covered_receipts:
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Automation draft PR proposals require candidate target component evidence (`ansa_components.candidate_target_components`).

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: rework-3fb31b1061ab
failure_class: rework
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - rework-3fb31b1061ab
  failure_classes:
    - rework
  target_surfaces:
    - library/templates/portfolio-repo-registry.yml
  covered_receipts:
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: rework-3fb31b1061ab
failure_class: rework
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - rework-3fb31b1061ab
  failure_classes:
    - rework
  target_surfaces:
    - library/templates/portfolio-repo-registry.yml
  covered_receipts:
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
scope_creep-d64f4b08cb35 blocked

Approve the issue proposal if you want a bounded follow-up recorded for `realms/coding/workflows/implement-issue/SKILL.md`; otherwise defer or waive.

  • Problem: Repeated evidence says scope creep is still causing follow-through friction around `realms/coding/workflows/implement-issue/SKILL.md`.
  • Follow-up: Record a bounded GitHub issue for `realms/coding/workflows/implement-issue/SKILL.md` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: scope_creep-d64f4b08cb35
failure_class: scope_creep
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: realms/coding/workflows/implement-issue/SKILL.md
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60scope_creep%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60scope_creep-d64f4b08cb35%60%0A-%20Evidence%20count%3A%2037%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20639%2C%20611%2C%20433%2C%20407%2C%20405%2C%20403%2C%20392%2C%20387%2C%20357%2C%20356%2C%20316%2C%20260%2C%20241%2C%20211%2C%20205%2C%20195%2C%20188%2C%20184%2C%20175%2C%20179%2C%20161%2C%20156%2C%20146%2C%20145%2C%20139%2C%20136%2C%20116%2C%20106%2C%20104%0A-%20Affected%20receipts%3A%20%60runs%2F2026-07-02-issue-692-afk-pr-settlement-contracts.yml%60%2C%20%60runs%2F2026-06-25-issue-630-review-surface-readback.yml%60%2C%20%60runs%2F2026-06-25-issue-633-session-learning-review.yml%60%2C%20%60runs%2F2026-06-24-all-prs-evidence-console.yml%60%2C%20%60runs%2F2026-06-24%2Fissue-607-enki-entrypoint-evidence%2Frun-receipt.yml%60%2C%20%60runs%2F2026-06-18-issue-427-external-skill-port.yml%60%2C%20%60runs%2F2026-06-17-issue-428-codex-desktop-compliance.yml%60%2C%20%60runs%2F2026-06-15-issue-406-setup-pre-commit.yml%60%2C%20%60runs%2F2026-06-15-issue-404-productivity-skills.yml%60%2C%20%60runs%2F2026-06-14-issue-400-architecture-review.yml%60%2C%20%60runs%2F2026-06-14-issue-401-planning-depth-modes.yml%60%2C%20%60runs%2F2026-06-14-issue-391-brain2-portfolio-memory.yml%60%2C%20%60runs%2F2026-06-13-issue-383-grill-with-docs.yml%60%2C%20%60runs%2F2026-06-08-issues-339-341-332-workflow-control.yml%60%2C%20%60runs%2F2026-06-08-issue-355-prompt-front-door.yml%60%2C%20%60runs%2F2026-06-03-issue-322-owner-transfer-cleanup.yml%60%2C%20%60runs%2F2026-06-02-issue-302-opencode-requesty-routing.yml%60%2C%20%60runs%2F2026-05-24-issue-255-finding-lifecycle-contract.yml%60%2C%20%60runs%2F2026-05-23-issue-227-operator-review-template-metadata.yml%60%2C%20%60runs%2F2026-05-22-issue-217-clean-baseline.yml%60%2C%20%60runs%2F2026-05-21-issue-210-self-improvement-candidate-escalation.yml%60%2C%20%60runs%2F2026-05-20-issue-201-opencode-staged-agent-pack.yml%60%2C%20%60runs%2F2026-05-19-issue-191-agent-feedback-event.yml%60%2C%20%60runs%2F2026-05-19-issue-172-173-cross-repo-self-improvement-boundary.yml%60%2C%20%60runs%2F2026-05-19-issue-174-cli-worker-adapters.yml%60%2C%20%60runs%2F2026-05-18-issue-167-114-prompt-coverage-proof-prompts.yml%60%2C%20%60runs%2F2026-05-18-issue-162-cross-repo-tranche-supervisor.yml%60%2C%20%60runs%2F2026-05-18-issue-157-158-160-preflight-fallback-opencode.yml%60%2C%20%60runs%2F2026-05-18-issue-152-154-pr-body-validation-closeout.yml%60%2C%20%60runs%2F2026-05-17-issue-93-transcription-session-workflows.yml%60%2C%20%60runs%2F2026-05-17-issue-143-allowance-first-policy.yml%60%2C%20%60runs%2F2026-05-17-issue-134-stage-model-routing.yml%60%2C%20%60runs%2F2026-05-16-issue-127-final-state-verification.yml%60%2C%20%60runs%2F2026-05-16-issue-119-opencode-adapter.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%2C%20%60runs%2F2026-05-14-issue-78-impact-scorecard.yml%60%2C%20%60runs%2F2026-05-14-issue-77-outcome-evidence.yml%60%0A-%20Affected%20components%3A%20%60realms%2Fcoding%2Fworkflows%2Fimplement-issue%2FSKILL.md%60%2C%20%60coding.implement-issue%60%2C%20%60library%2Fadapters%2Fmodel-routing-standard.md%60%2C%20%60library%2Ftemplates%2Fcompact-stage-handoff.md%60%2C%20%60library%2Ftemplates%2Ffinding-lifecycle-contract.yml%60%2C%20%60library%2Ftemplates%2Fgithub-review-surface-readback.yml%60%2C%20%60library%2Ftemplates%2Fplanning-evidence-table.md%60%2C%20%60library%2Ftemplates%2Ftranscript-decomposition.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fattack-review%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fautonomous-pr-cycle%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fresolve-merge-rebase-conflicts%2FSKILL.md%60%2C%20%60realms%2Foperations%2Fworkflows%2Fagent-feedback-review%2FSKILL.md%60%2C%20%60realms%2Foperations%2Fworkflows%2Fsession-learning-review%2FSKILL.md%60%2C%20%60realms%2Fproductivity%2Fworkflows%2Fteach%2FSKILL.md%60%2C%20%60scripts%2Fcore%2Fgenerate-ansa-scorecard.mjs%60%2C%20%60scripts%2Fcore%2Fgenerate-evidence-console-dashboard.mjs%60%2C%20%60scripts%2Fcore%2Fvalidate-library.mjs%60%0A-%20New%20component%20needed%3A%20true%0A-%20Proposed%20target%20surface%3A%20%60realms%2Fcoding%2Fworkflows%2Fimplement-issue%2FSKILL.md%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%2C%20%60ansa_components.workflows_used%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%2037%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60scope_creep%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - scope_creep-d64f4b08cb35
  failure_classes:
    - scope_creep
  target_surfaces:
    - realms/coding/workflows/implement-issue/SKILL.md
  covered_receipts:
    - runs/2026-07-02-issue-692-afk-pr-settlement-contracts.yml
    - runs/2026-06-25-issue-630-review-surface-readback.yml
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-18-issue-427-external-skill-port.yml
    - runs/2026-06-17-issue-428-codex-desktop-compliance.yml
    - runs/2026-06-15-issue-406-setup-pre-commit.yml
    - runs/2026-06-15-issue-404-productivity-skills.yml
    - runs/2026-06-14-issue-400-architecture-review.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-14-issue-391-brain2-portfolio-memory.yml
    - runs/2026-06-13-issue-383-grill-with-docs.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-08-issue-355-prompt-front-door.yml
    - runs/2026-06-03-issue-322-owner-transfer-cleanup.yml
    - runs/2026-06-02-issue-302-opencode-requesty-routing.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-217-clean-baseline.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-20-issue-201-opencode-staged-agent-pack.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-19-issue-172-173-cross-repo-self-improvement-boundary.yml
    - runs/2026-05-19-issue-174-cli-worker-adapters.yml
    - runs/2026-05-18-issue-167-114-prompt-coverage-proof-prompts.yml
    - runs/2026-05-18-issue-162-cross-repo-tranche-supervisor.yml
    - runs/2026-05-18-issue-157-158-160-preflight-fallback-opencode.yml
    - runs/2026-05-18-issue-152-154-pr-body-validation-closeout.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-143-allowance-first-policy.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-16-issue-127-final-state-verification.yml
    - runs/2026-05-16-issue-119-opencode-adapter.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Automation draft PR proposals require candidate target component evidence (`ansa_components.candidate_target_components`).

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: scope_creep-d64f4b08cb35
failure_class: scope_creep
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - scope_creep-d64f4b08cb35
  failure_classes:
    - scope_creep
  target_surfaces:
    - realms/coding/workflows/implement-issue/SKILL.md
  covered_receipts:
    - runs/2026-07-02-issue-692-afk-pr-settlement-contracts.yml
    - runs/2026-06-25-issue-630-review-surface-readback.yml
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-18-issue-427-external-skill-port.yml
    - runs/2026-06-17-issue-428-codex-desktop-compliance.yml
    - runs/2026-06-15-issue-406-setup-pre-commit.yml
    - runs/2026-06-15-issue-404-productivity-skills.yml
    - runs/2026-06-14-issue-400-architecture-review.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-14-issue-391-brain2-portfolio-memory.yml
    - runs/2026-06-13-issue-383-grill-with-docs.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-08-issue-355-prompt-front-door.yml
    - runs/2026-06-03-issue-322-owner-transfer-cleanup.yml
    - runs/2026-06-02-issue-302-opencode-requesty-routing.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-217-clean-baseline.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-20-issue-201-opencode-staged-agent-pack.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-19-issue-172-173-cross-repo-self-improvement-boundary.yml
    - runs/2026-05-19-issue-174-cli-worker-adapters.yml
    - runs/2026-05-18-issue-167-114-prompt-coverage-proof-prompts.yml
    - runs/2026-05-18-issue-162-cross-repo-tranche-supervisor.yml
    - runs/2026-05-18-issue-157-158-160-preflight-fallback-opencode.yml
    - runs/2026-05-18-issue-152-154-pr-body-validation-closeout.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-143-allowance-first-policy.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-16-issue-127-final-state-verification.yml
    - runs/2026-05-16-issue-119-opencode-adapter.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: scope_creep-d64f4b08cb35
failure_class: scope_creep
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - scope_creep-d64f4b08cb35
  failure_classes:
    - scope_creep
  target_surfaces:
    - realms/coding/workflows/implement-issue/SKILL.md
  covered_receipts:
    - runs/2026-07-02-issue-692-afk-pr-settlement-contracts.yml
    - runs/2026-06-25-issue-630-review-surface-readback.yml
    - runs/2026-06-25-issue-633-session-learning-review.yml
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-24/issue-607-enki-entrypoint-evidence/run-receipt.yml
    - runs/2026-06-18-issue-427-external-skill-port.yml
    - runs/2026-06-17-issue-428-codex-desktop-compliance.yml
    - runs/2026-06-15-issue-406-setup-pre-commit.yml
    - runs/2026-06-15-issue-404-productivity-skills.yml
    - runs/2026-06-14-issue-400-architecture-review.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-06-14-issue-391-brain2-portfolio-memory.yml
    - runs/2026-06-13-issue-383-grill-with-docs.yml
    - runs/2026-06-08-issues-339-341-332-workflow-control.yml
    - runs/2026-06-08-issue-355-prompt-front-door.yml
    - runs/2026-06-03-issue-322-owner-transfer-cleanup.yml
    - runs/2026-06-02-issue-302-opencode-requesty-routing.yml
    - runs/2026-05-24-issue-255-finding-lifecycle-contract.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-217-clean-baseline.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-20-issue-201-opencode-staged-agent-pack.yml
    - runs/2026-05-19-issue-191-agent-feedback-event.yml
    - runs/2026-05-19-issue-172-173-cross-repo-self-improvement-boundary.yml
    - runs/2026-05-19-issue-174-cli-worker-adapters.yml
    - runs/2026-05-18-issue-167-114-prompt-coverage-proof-prompts.yml
    - runs/2026-05-18-issue-162-cross-repo-tranche-supervisor.yml
    - runs/2026-05-18-issue-157-158-160-preflight-fallback-opencode.yml
    - runs/2026-05-18-issue-152-154-pr-body-validation-closeout.yml
    - runs/2026-05-17-issue-93-transcription-session-workflows.yml
    - runs/2026-05-17-issue-143-allowance-first-policy.yml
    - runs/2026-05-17-issue-134-stage-model-routing.yml
    - runs/2026-05-16-issue-127-final-state-verification.yml
    - runs/2026-05-16-issue-119-opencode-adapter.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
sequencing_problem-b762366e22bb blocked

Approve the issue proposal if you want a bounded follow-up recorded for `coding.autonomous-pr-cycle`; otherwise defer or waive.

  • Problem: Repeated evidence says sequencing problem is still causing follow-through friction around `coding.autonomous-pr-cycle`.
  • Follow-up: Record a bounded GitHub issue for `coding.autonomous-pr-cycle` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: sequencing_problem-b762366e22bb
failure_class: sequencing_problem
issue_title: Amend existing corrective issue with repeated self-improvement evidence
target_surface: coding.autonomous-pr-cycle
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Amend%20existing%20corrective%20issue%20with%20repeated%20self-improvement%20evidence&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60sequencing_problem%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60sequencing_problem-b762366e22bb%60%0A-%20Evidence%20count%3A%2010%0A-%20Consecutive%20evidence%20count%3A%201%0A-%20Affected%20PRs%3A%20741%2C%20534%2C%20351%2C%20242%2C%20241%2C%20240%0A-%20Affected%20receipts%3A%20%60runs%2F2026-07-07-issue-734-pr-body-preflight.yml%60%2C%20%60runs%2F2026-06-25-issue-628-red-team-lens-completeness.yml%60%2C%20%60runs%2F2026-06-20-issue-531-pr-tranche-closeout.yml%60%2C%20%60runs%2F2026-06-07-reusable-conformance-explicit-token-checkout.yml%60%2C%20%60runs%2F2026-05-23-issue-228-stale-pr-sweeper-report.yml%60%2C%20%60runs%2F2026-05-23-issue-227-operator-review-template-metadata.yml%60%2C%20%60runs%2F2026-05-22-issue-222-operator-decision-review-pack.yml%60%2C%20%60runs%2F2026-05-22-issue-225-tranche-manifest-active-held-lanes.yml%60%2C%20%60runs%2F2026-05-22-issue-232-internal-source-reuse-policy.yml%60%2C%20%60runs%2F2026-05-19-issue-180-sticky-evidence-hard-gates.yml%60%0A-%20Affected%20components%3A%20%60coding.autonomous-pr-cycle%60%2C%20%60realms%2Fcoding%2Frun-desks%2Fpr-tranche-closeout%2FREADME.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fattack-review%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fautonomous-pr-cycle%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Ffinalise-pr%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fimplement-issue%2FSKILL.md%60%2C%20%60realms%2Fcoding%2Fworkflows%2Fverify-pr%2FSKILL.md%60%2C%20%60realms%2Fproductivity%2Fworkflows%2Fred-team%2FSKILL.md%60%2C%20%60scripts%2Fcore%2Fvalidate-pr-body.mjs%60%0A-%20New%20component%20needed%3A%20true%0A-%20Proposed%20target%20surface%3A%20%60coding.autonomous-pr-cycle%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%2010%20matching%20receipts%20and%201%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60sequencing_problem%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Attach%20the%20repeated%20evidence%20to%20the%20existing%20issue%20instead%20of%20opening%20a%20duplicate.%0A-%20Keep%20the%20amendment%20proposal-only%20until%20a%20human%20selects%20the%20owning%20issue.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - sequencing_problem-b762366e22bb
  failure_classes:
    - sequencing_problem
  target_surfaces:
    - coding.autonomous-pr-cycle
  covered_receipts:
    - runs/2026-07-07-issue-734-pr-body-preflight.yml
    - runs/2026-06-25-issue-628-red-team-lens-completeness.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-07-reusable-conformance-explicit-token-checkout.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-22-issue-225-tranche-manifest-active-held-lanes.yml
    - runs/2026-05-22-issue-232-internal-source-reuse-policy.yml
    - runs/2026-05-19-issue-180-sticky-evidence-hard-gates.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Recommended action amend_existing_issue requires human review instead of automation draft PR generation.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: sequencing_problem-b762366e22bb
failure_class: sequencing_problem
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - sequencing_problem-b762366e22bb
  failure_classes:
    - sequencing_problem
  target_surfaces:
    - coding.autonomous-pr-cycle
  covered_receipts:
    - runs/2026-07-07-issue-734-pr-body-preflight.yml
    - runs/2026-06-25-issue-628-red-team-lens-completeness.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-07-reusable-conformance-explicit-token-checkout.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-22-issue-225-tranche-manifest-active-held-lanes.yml
    - runs/2026-05-22-issue-232-internal-source-reuse-policy.yml
    - runs/2026-05-19-issue-180-sticky-evidence-hard-gates.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: sequencing_problem-b762366e22bb
failure_class: sequencing_problem
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - sequencing_problem-b762366e22bb
  failure_classes:
    - sequencing_problem
  target_surfaces:
    - coding.autonomous-pr-cycle
  covered_receipts:
    - runs/2026-07-07-issue-734-pr-body-preflight.yml
    - runs/2026-06-25-issue-628-red-team-lens-completeness.yml
    - runs/2026-06-20-issue-531-pr-tranche-closeout.yml
    - runs/2026-06-07-reusable-conformance-explicit-token-checkout.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-22-issue-225-tranche-manifest-active-held-lanes.yml
    - runs/2026-05-22-issue-232-internal-source-reuse-policy.yml
    - runs/2026-05-19-issue-180-sticky-evidence-hard-gates.yml
  reason: <fill in operator-approved reason>
stale_canon-4859578a9a3e blocked

Approve the issue proposal if you want a bounded follow-up recorded for `realms/coding/workflows/finalise-pr/SKILL.md`; otherwise defer or waive.

  • Problem: Repeated evidence says stale canon is still causing follow-through friction around `realms/coding/workflows/finalise-pr/SKILL.md`.
  • Follow-up: Record a bounded GitHub issue for `realms/coding/workflows/finalise-pr/SKILL.md` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: stale_canon-4859578a9a3e
failure_class: stale_canon
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: realms/coding/workflows/finalise-pr/SKILL.md
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60stale_canon%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60stale_canon-4859578a9a3e%60%0A-%20Evidence%20count%3A%2012%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20611%2C%20437%2C%20440%2C%20257%2C%20242%2C%20241%2C%20240%2C%20212%2C%20211%2C%20208%2C%20138%0A-%20Affected%20receipts%3A%20%60runs%2F2026-06-24-all-prs-evidence-console.yml%60%2C%20%60runs%2F2026-06-18-pr-437-self-improvement-refresh.yml%60%2C%20%60runs%2F2026-06-18-pr-440-self-improvement-refresh.yml%60%2C%20%60runs%2F2026-06-14-issue-401-planning-depth-modes.yml%60%2C%20%60runs%2F2026-05-24-issue-253-portfolio-registry-blocked.yml%60%2C%20%60runs%2F2026-05-23-issue-228-stale-pr-sweeper-report.yml%60%2C%20%60runs%2F2026-05-23-issue-227-operator-review-template-metadata.yml%60%2C%20%60runs%2F2026-05-22-issue-222-operator-decision-review-pack.yml%60%2C%20%60runs%2F2026-05-21-pr-212-self-improvement-refresh-closeout.yml%60%2C%20%60runs%2F2026-05-21-issue-210-self-improvement-candidate-escalation.yml%60%2C%20%60runs%2F2026-05-21-issue-200-review-thread-zero-gate.yml%60%2C%20%60runs%2F2026-05-17-issue-133-batch-closeout-report.yml%60%0A-%20Affected%20components%3A%20%60realms%2Fcoding%2Fworkflows%2Ffinalise-pr%2FSKILL.md%60%2C%20%60library%2Ftemplates%2Fbatch-closeout-report.md%60%2C%20%60library%2Ftemplates%2Fplanning-evidence-table.md%60%2C%20%60library%2Ftemplates%2Fportfolio-repo-registry.yml%60%2C%20%60scripts%2Fcore%2Fgenerate-ansa-scorecard.mjs%60%2C%20%60scripts%2Fcore%2Fgenerate-evidence-console-dashboard.mjs%60%2C%20%60scripts%2Fcore%2Fvalidate-library.mjs%60%0A-%20New%20component%20needed%3A%20true%0A-%20Proposed%20target%20surface%3A%20%60realms%2Fcoding%2Fworkflows%2Ffinalise-pr%2FSKILL.md%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%2012%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60stale_canon%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - stale_canon-4859578a9a3e
  failure_classes:
    - stale_canon
  target_surfaces:
    - realms/coding/workflows/finalise-pr/SKILL.md
  covered_receipts:
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-18-pr-437-self-improvement-refresh.yml
    - runs/2026-06-18-pr-440-self-improvement-refresh.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-21-pr-212-self-improvement-refresh-closeout.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-21-issue-200-review-thread-zero-gate.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Failure class is not allowlisted for automation draft PR proposals.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: stale_canon-4859578a9a3e
failure_class: stale_canon
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - stale_canon-4859578a9a3e
  failure_classes:
    - stale_canon
  target_surfaces:
    - realms/coding/workflows/finalise-pr/SKILL.md
  covered_receipts:
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-18-pr-437-self-improvement-refresh.yml
    - runs/2026-06-18-pr-440-self-improvement-refresh.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-21-pr-212-self-improvement-refresh-closeout.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-21-issue-200-review-thread-zero-gate.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: stale_canon-4859578a9a3e
failure_class: stale_canon
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - stale_canon-4859578a9a3e
  failure_classes:
    - stale_canon
  target_surfaces:
    - realms/coding/workflows/finalise-pr/SKILL.md
  covered_receipts:
    - runs/2026-06-24-all-prs-evidence-console.yml
    - runs/2026-06-18-pr-437-self-improvement-refresh.yml
    - runs/2026-06-18-pr-440-self-improvement-refresh.yml
    - runs/2026-06-14-issue-401-planning-depth-modes.yml
    - runs/2026-05-24-issue-253-portfolio-registry-blocked.yml
    - runs/2026-05-23-issue-228-stale-pr-sweeper-report.yml
    - runs/2026-05-23-issue-227-operator-review-template-metadata.yml
    - runs/2026-05-22-issue-222-operator-decision-review-pack.yml
    - runs/2026-05-21-pr-212-self-improvement-refresh-closeout.yml
    - runs/2026-05-21-issue-210-self-improvement-candidate-escalation.yml
    - runs/2026-05-21-issue-200-review-thread-zero-gate.yml
    - runs/2026-05-17-issue-133-batch-closeout-report.yml
  reason: <fill in operator-approved reason>
weak_scope-7b6b235e81d7 blocked

Approve the issue proposal if you want a bounded follow-up recorded for `unknown`; otherwise defer or waive.

  • Problem: Repeated evidence says weak scope is still causing follow-through friction around `unknown`.
  • Follow-up: Record a bounded GitHub issue for `unknown` so the improvement is durable repo work instead of a chat-only decision.
  • Approve: Approval is not complete when the text is pasted; an agent must create or update the GitHub issue/PR evidence and add a reviewed resolution entry.
  • Defer: Deferring keeps the candidate visible and records when or why it should be revisited.
  • Suppress: Waiving suppresses this candidate only for the recorded scope and reason; fresh uncovered evidence can still surface later.
  • Durable state: Durable state lives in `docs/audits/self-improvement-candidates/resolutions.yml` plus the linked GitHub issue, PR, or review-thread evidence.

Do not treat the refresh PR as routine merge-ready until one of these actions is chosen and recorded.

Open issue available medium

Approve issue proposal

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Creates or updates a GitHub issue and records the candidate as resolved against that issue.
Required state
Open the issue URL, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate stops blocking only after the issue/update and resolution entry are recorded.
Issue URL
Open issue form

none

APPROVE ISSUE PROPOSAL
candidate_id: weak_scope-7b6b235e81d7
failure_class: weak_scope
issue_title: Tighten workflow, template, or adapter guidance for repeated self-improvement misses
target_surface: unknown
github_issue_url: https://github.com/Eunoetic/ansa/issues/new?title=Tighten%20workflow%2C%20template%2C%20or%20adapter%20guidance%20for%20repeated%20self-improvement%20misses&body=%23%23%20Summary%0ARepeated%20self-improvement%20evidence%20surfaced%20failure%20class%20%60weak_scope%60.%0A%0A%23%23%20Evidence%0A-%20Candidate%20id%3A%20%60weak_scope-7b6b235e81d7%60%0A-%20Evidence%20count%3A%206%0A-%20Consecutive%20evidence%20count%3A%200%0A-%20Affected%20PRs%3A%20348%2C%20187%2C%20166%2C%20116%2C%20106%2C%20104%0A-%20Affected%20receipts%3A%20%60runs%2F2026-06-07-issue-337-pr-body-contract.yml%60%2C%20%60runs%2F2026-05-19-issue-10-minimal-agent-pack-export.yml%60%2C%20%60runs%2F2026-05-18-issue-164-165-self-improvement-loop-automation.yml%60%2C%20%60runs%2F2026-05-15-issue-111-sticky-pr-comment.yml%60%2C%20%60runs%2F2026-05-14-issue-78-impact-scorecard.yml%60%2C%20%60runs%2F2026-05-14-issue-77-outcome-evidence.yml%60%0A-%20Affected%20components%3A%20unknown%0A-%20New%20component%20needed%3A%20false%0A-%20Proposed%20target%20surface%3A%20%60unknown%60%0A-%20Missing%20receipt%20fields%3A%20%60ansa_components.candidate_target_components%60%0A%0A%23%23%20Repeated%20Evidence%0AThis%20candidate%20is%20based%20on%206%20matching%20receipts%20and%200%20consecutive%20matching%20receipts%20in%20the%20selected%20evidence%20window.%0A%0A%23%23%20Non-Goals%0A-%20Do%20not%20auto-create%20a%20GitHub%20issue%20or%20implementation%20PR%20from%20this%20candidate.%0A-%20Do%20not%20broaden%20the%20remediation%20beyond%20the%20named%20failure%20class%20and%20target%20surface.%0A%0A%23%23%20Acceptance%20Criteria%0A-%20Keep%20the%20remediation%20scoped%20to%20the%20single%20failure%20class%20%60weak_scope%60.%0A-%20Preserve%20dry-run-first%20behavior%20and%20deterministic%20output%20paths.%0A-%20Do%20not%20grant%20merge-ready%20status%20without%20hosted%20checks%20or%20explicit%20manual%20validation%20evidence.%0A-%20Update%20the%20smallest%20exact%20workflow%2C%20template%2C%20adapter%2C%20or%20proposed%20inactive%20component%20named%20by%20repeated%20receipt%20evidence.%0A-%20Keep%20repo-specific%20policy%20in%20adapters%20and%20add%20or%20refresh%20tests%20if%20validation%20behavior%20changes.
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: create_or_update_github_issue_then_record_resolution
decision_reason: approved bounded follow-up issue
resolution_entry:
- resolution_id: <fill in stable id>
  status: resolved
  reference: <created or updated GitHub issue URL>
  candidate_ids:
    - weak_scope-7b6b235e81d7
  failure_classes:
    - weak_scope
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Approve unavailable high

Approve draft PR proposal

Kind
blocked_action
Copy role
blocked_reason
Durable effect
Authorises a bounded implementation PR lane and links the candidate to that lane.
Required state
Open the draft PR only when policy allows it, then add a reviewed resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml`.
After action
The candidate remains active until the draft PR exists and the resolution entry links to it.
Issue URL
none

Automation draft PR proposals require a concrete target surface and candidate target component evidence.

Defer available low

Defer candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Keeps the candidate visible with a concrete revisit trigger instead of silently dropping it.
Required state
Add a deferred resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with a revisit trigger.
After action
The candidate can resurface when the revisit trigger or new evidence applies.
Issue URL
none

none

DEFER CANDIDATE
candidate_id: weak_scope-7b6b235e81d7
failure_class: weak_scope
decision_reason: <fill in deferral reason>
revisit_trigger: <fill in date or condition>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_deferred_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: deferred
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - weak_scope-7b6b235e81d7
  failure_classes:
    - weak_scope
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>
Suppress available medium

Waive candidate

Kind
durable_repo_update
Copy role
repo_update_instruction
Durable effect
Suppresses the candidate only for the explicit waiver scope and reason.
Required state
Add a waived resolution entry to `docs/audits/self-improvement-candidates/resolutions.yml` with scope and reason.
After action
Fresh uncovered evidence can still create a new candidate after the waiver.
Issue URL
none

none

WAIVE CANDIDATE
candidate_id: weak_scope-7b6b235e81d7
failure_class: weak_scope
decision_reason: <fill in waiver reason>
waiver_scope: <fill in scope>
resolution_file: docs/audits/self-improvement-candidates/resolutions.yml
durable_action_required: record_waived_resolution_entry
resolution_entry:
- resolution_id: <fill in stable id>
  status: waived
  reference: <PR comment, issue comment, or operator decision URL>
  candidate_ids:
    - weak_scope-7b6b235e81d7
  failure_classes:
    - weak_scope
  target_surfaces:
    - unknown
  covered_receipts:
    - runs/2026-06-07-issue-337-pr-body-contract.yml
    - runs/2026-05-19-issue-10-minimal-agent-pack-export.yml
    - runs/2026-05-18-issue-164-165-self-improvement-loop-automation.yml
    - runs/2026-05-15-issue-111-sticky-pr-comment.yml
    - runs/2026-05-14-issue-78-impact-scorecard.yml
    - runs/2026-05-14-issue-77-outcome-evidence.yml
  reason: <fill in operator-approved reason>

Resolved Candidates

adapter_limitation-6ecc61a1d8aa adapter_limitation realms/coding/workflows/finalise-pr/SKILL.md 44 resolved https://github.com/Eunoetic/ansa/issues/675 Operator approved follow-up issue #675 for the repeated finalise-pr adapter-limitation evidence surfaced by PR #656.
late_review_check_settling-c16646a78432 late_review_check_settling realms/productivity/workflows/red-team/SKILL.md 3 resolved https://github.com/Eunoetic/ansa/issues/675 Operator approved tracking the late review/check-settling candidate against issue #675 so the improvement is preserved as normal issue-backed Ansa work instead of remaining only a PR comment or chat decision.

Chains And Workbench

derived Derived only from repeated outcome-learning evidence that already names route chains.

Route-chain candidates: 4 | Workbench candidates: 0

Route Invocation Hints

business-development1 routes
business-development.opportunity-briefing read_only business-development bounded opportunity briefing or outreach-preparation summary not_available realms/business-development/workflows/opportunity-briefing/SKILL.md
cartant9 routes
cartant.contact-status-sweep mixed cartant bounded contact-status sweep with follow-up evidence, restriction flags, and approval-gated proposed updates library/prompts/route-stubs/cartant-contact-status-sweep.md realms/cartant/workflows/contact-status-sweep/SKILL.md
cartant.gmail-job-label-triage read_only cartant metadata-first Gmail job-label triage that recommends one safest next thread or item library/prompts/route-stubs/cartant-gmail-job-label-triage.md realms/cartant/workflows/gmail-job-label-triage/SKILL.md
cartant.meeting-transcript-processing mixed cartant bounded meeting transcript processing into reviewable tracker-side follow-up library/prompts/route-stubs/cartant-meeting-transcript-processing.md realms/cartant/workflows/meeting-transcript-processing/SKILL.md
cartant.pre-meeting-brief read_only cartant bounded pre-meeting brief from current tracker evidence plus selected support material library/prompts/route-stubs/cartant-pre-meeting-brief.md realms/cartant/workflows/pre-meeting-brief/SKILL.md
cartant.start-day read_only cartant read-only Cartant start-day or session-opening orientation that routes to the smallest safe next workflow library/prompts/route-stubs/cartant-start-day.md realms/cartant/workflows/start-day/SKILL.md
cartant.strategic-next-step-review mixed cartant bounded strategic next-step review with optional dry-run tracker updates library/prompts/route-stubs/cartant-strategic-next-step-review.md realms/cartant/workflows/strategic-next-step-review/SKILL.md
cartant.targeted-message-processing mixed cartant exact one-source-item processing from Gmail or bounded pasted evidence into evidence-backed proposed tracker writes library/prompts/route-stubs/cartant-targeted-message-processing.md realms/cartant/workflows/targeted-message-processing/SKILL.md
cartant.tracker-hygiene mixed cartant bounded tracker-hygiene subflows with approval-gated proposed writes library/prompts/route-stubs/cartant-tracker-hygiene.md realms/cartant/workflows/tracker-hygiene/SKILL.md
cartant.voice-note-processing mixed cartant bounded voice-note processing into reviewable tracker-side follow-up library/prompts/route-stubs/cartant-voice-note-processing.md realms/cartant/workflows/voice-note-processing/SKILL.md
coding28 routes
coding.afk-goal-writer read_only coding write bounded AFK coding goals before unattended or semi-unattended agent work library/prompts/route-stubs/coding-afk-goal-writer.md realms/coding/workflows/afk-goal-writer/SKILL.md
coding.agent-run-launch-kit read_only coding produce one terminal setup block plus one supervisor prompt for session, worktree, and isolated coding lanes library/prompts/route-stubs/coding-agent-run-launch-kit.md realms/coding/workflows/agent-run-launch-kit/SKILL.md
coding.ansa-operator-prompt read_only coding generate or slim a pasteable Ansa operator prompt by inferring an existing coding route or route chain library/prompts/route-stubs/coding-ansa-operator-prompt.md realms/coding/workflows/ansa-operator-prompt/SKILL.md
coding.architecture-review mixed coding report-first architecture review that finds deepening opportunities through bounded repo and docs exploration library/prompts/route-stubs/coding-architecture-review.md realms/coding/workflows/architecture-review/SKILL.md
coding.attack-review read_only coding adversarial review or red-team pass library/prompts/route-stubs/coding-attack-review.md realms/coding/workflows/attack-review/SKILL.md
coding.autonomous-pr-cycle mixed coding full autonomous issue-to-PR coding cycle library/prompts/route-stubs/coding-autonomous-pr-cycle.md realms/coding/workflows/autonomous-pr-cycle/SKILL.md
coding.batch-run-planner read_only coding plan safe parallel issue and PR tranches with sequential gates and merge sequencing library/prompts/route-stubs/coding-batch-run-planner.md realms/coding/workflows/batch-run-planner/SKILL.md
coding.clean-baseline mixed coding dry-run-first post-tranche session-surface, worktree, local-branch, and repo-baseline cleanup with backups and receipts library/prompts/route-stubs/coding-clean-baseline.md realms/coding/workflows/clean-baseline/SKILL.md
coding.finalise-pr read_only coding final pre-merge closeout and merge sequencing library/prompts/route-stubs/coding-finalise-pr.md realms/coding/workflows/finalise-pr/SKILL.md
coding.goose-controller-proof-prompt read_only coding generate a bounded Goose controller-proof prompt with evidence-first proof gates library/prompts/route-stubs/coding-goose-controller-proof-prompt.md realms/coding/workflows/goose-controller-proof-prompt/SKILL.md
coding.grill-with-docs mixed coding docs-aware plan challenge that sharpens terminology and decision records through target repo canon and adapters library/prompts/route-stubs/coding-grill-with-docs.md realms/coding/workflows/grill-with-docs/SKILL.md
coding.implement-issue mutating coding authorised issue implementation or fix pass library/prompts/route-stubs/coding-implement-issue.md realms/coding/workflows/implement-issue/SKILL.md
coding.plan-issue read_only coding issue planning before implementation library/prompts/route-stubs/coding-issue-plan.md realms/coding/workflows/plan-issue/SKILL.md
coding.pr-review read_only coding bounded PR review or stage-aware tranche review routing library/prompts/route-stubs/coding-pr-review.md realms/coding/workflows/pr-review/SKILL.md
coding.prototype mixed coding answer one runnable design question with a disposable non-production experiment and findings handoff library/prompts/route-stubs/coding-prototype.md realms/coding/workflows/prototype/SKILL.md
coding.red-team-fix-prompt read_only coding generate a bounded red-team review or fix-loop prompt that reuses existing coding review stages library/prompts/route-stubs/coding-red-team-review-and-fix.md realms/coding/workflows/red-team-fix-prompt/SKILL.md
coding.repo-hygiene-cleanup mixed coding dry-run-first local branch and worktree cleanup after agent or pull request work not_available realms/coding/workflows/repo-hygiene-cleanup/SKILL.md
coding.research mixed coding bounded primary-source technical research persisted as one cited repository Markdown note when authorised library/prompts/route-stubs/coding-research.md realms/coding/workflows/research/SKILL.md
coding.resolve-merge-rebase-conflicts mixed coding resolve in-progress merge, rebase, cherry-pick, or patch conflicts by inspecting source intent and validating the result library/prompts/route-stubs/coding-resolve-merge-rebase-conflicts.md realms/coding/workflows/resolve-merge-rebase-conflicts/SKILL.md
coding.review-comment-closeout mixed coding close out review comments or unresolved threads library/prompts/route-stubs/coding-review-comment-closeout.md realms/coding/workflows/review-comment-closeout/SKILL.md
coding.safe-branch-main-reset mixed coding safely switch or reset back to the adapter-defined default branch after verifying repo state not_available realms/coding/workflows/safe-branch-main-reset/SKILL.md
coding.setup-pre-commit mixed coding package-manager-aware Node pre-commit hook setup with explicit write and install gates library/prompts/route-stubs/coding-setup-pre-commit.md realms/coding/workflows/setup-pre-commit/SKILL.md
coding.sync-local-repos-main mixed coding dry-run-first multi-repo default-branch sync with stashing and branch-preserving fetches library/prompts/route-stubs/coding-sync-local-repos-main.md realms/coding/workflows/sync-local-repos-main/SKILL.md
coding.to-spec mixed coding synthesise settled context into one target-repository-owned buildable specification without reopening decisions library/prompts/route-stubs/coding-to-spec.md realms/coding/workflows/to-spec/SKILL.md
coding.to-tickets mixed coding draft deduped dependency-aware vertical implementation tickets from one accepted specification library/prompts/route-stubs/coding-to-tickets.md realms/coding/workflows/to-tickets/SKILL.md
coding.triage read_only coding classify one incoming work item from verified evidence and recommend the smallest safe next route library/prompts/route-stubs/coding-triage.md realms/coding/workflows/triage/SKILL.md
coding.unattended-build-loop mixed coding provider-agnostic unattended plan-build-verify-harden-receipt contract not_available realms/coding/workflows/unattended-build-loop/SKILL.md
coding.verify-pr read_only coding PR readiness and validation evidence verification library/prompts/route-stubs/coding-verify-pr.md realms/coding/workflows/verify-pr/SKILL.md
household-shopping2 routes
household-shopping.child-clothing-shopping-plan read_only household-shopping plan child clothing and school-uniform shopping runs with policy interpretation, sizing, pack maths, and review-ready baskets not_available realms/household-shopping/workflows/child-clothing-shopping-plan/SKILL.md
household-shopping.shopping-agent-plan read_only household-shopping plan and execute secure household-shopping run under strict no-checkout invariants not_available realms/household-shopping/workflows/shopping-agent-plan/SKILL.md
operations12 routes
operations.agent-feedback-review mixed operations capture bounded diagnostic agent feedback events without creating a second receipt system library/prompts/route-stubs/operations-agent-feedback-review.md realms/operations/workflows/agent-feedback-review/SKILL.md
operations.ansa-self-maintenance mixed operations turn reviewed maintenance findings into bounded follow-up issues or proposals not_available realms/operations/workflows/ansa-self-maintenance/SKILL.md
operations.ansa-self-review read_only operations check whether implemented Ansa work achieved its intended outcome not_available realms/operations/workflows/ansa-self-review/SKILL.md
operations.brain2-portfolio-memory-proposed-write mixed operations approval-gated Brain 2 portfolio memory proposed-write packaging for cross-repo decisions and agreed language library/prompts/route-stubs/operations-brain2-portfolio-memory-proposed-write.md realms/operations/workflows/brain2-portfolio-memory-proposed-write/SKILL.md
operations.closed-issue-audit read_only operations audit whether closed issues are genuinely satisfied and evaluate their evidence chain not_available realms/operations/workflows/closed-issue-audit/SKILL.md
operations.create-issue-from-evidence mixed operations create or comment on GitHub issues from operator requests or bounded evidence after dedupe and explicit permission library/prompts/route-stubs/operations-create-issue-from-evidence.md realms/operations/workflows/create-issue-from-evidence/SKILL.md
operations.cross-repo-receipt-review read_only operations review target-repo receipts through bounded evidence without importing them as canon not_available realms/operations/workflows/cross-repo-receipt-review/SKILL.md
operations.outcome-learning-review read_only operations review outcome receipts and scorecards to classify repeated failures, bounded actions, and gate strength not_available realms/operations/workflows/outcome-learning-review/SKILL.md
operations.post-merge-bug-sweep read_only operations audit recent merged changes for regressions, bug signals, and evidence-backed follow-up routing not_available realms/operations/workflows/post-merge-bug-sweep/SKILL.md
operations.receipt-learning-review read_only operations review receipts and related evidence for reusable Ansa improvements not_available realms/operations/workflows/receipt-learning-review/SKILL.md
operations.session-learning-review read_only operations review a bounded Codex or coding-agent session for learning, waste, and follow-up candidates without writing by default library/prompts/route-stubs/operations-session-learning-review.md realms/operations/workflows/session-learning-review/SKILL.md
operations.skillopt-skill-pilot mixed operations run a bounded SkillOpt candidate-generation pilot against one Ansa skill surface with evidence-only outputs library/prompts/route-stubs/operations-skillopt-skill-pilot.md realms/operations/workflows/skillopt-skill-pilot/SKILL.md
planning2 routes
planning.event-planning read_only planning practical event planning for personal or work events from bounded evidence not_available realms/planning/workflows/event-planning/SKILL.md
planning.trip-planning read_only planning practical trip planning from capture-first notes and bounded evidence not_available realms/planning/workflows/trip-planning/SKILL.md
productivity12 routes
productivity.design-critique read_only productivity evidence-calibrated UI design critique across aesthetics, geometry, usability, accessibility, and practicality library/prompts/route-stubs/productivity-design-critique.md realms/productivity/workflows/design-critique/SKILL.md
productivity.gmail-inbox-management mixed productivity stateful Gmail inbox management with draft-only replies and approval-gated mailbox actions library/prompts/route-stubs/productivity-gmail-inbox-management.md realms/productivity/workflows/gmail-inbox-management/SKILL.md
productivity.grill-me read_only productivity one-question-at-a-time plan or design interrogation until shared understanding is reached library/prompts/route-stubs/productivity-grill-me.md realms/productivity/workflows/grill-me/SKILL.md
productivity.handoff mixed productivity compact redacted continuation packet for a future agent or session library/prompts/route-stubs/productivity-handoff.md realms/productivity/workflows/handoff/SKILL.md
productivity.legal-document-review-response mixed productivity practical legal-document review, issue spotting, source-grounded risk framing, and response drafting without legal advice claims library/prompts/route-stubs/productivity-legal-document-review-response.md realms/productivity/workflows/legal-document-review-response/SKILL.md
productivity.red-team read_only productivity red-team an idea using one or more code, product, and commercial lenses library/prompts/route-stubs/productivity-red-team.md realms/productivity/workflows/red-team/SKILL.md
productivity.teach mixed productivity stateful teaching workspace for learning a skill or concept over multiple sessions library/prompts/route-stubs/productivity-teach.md realms/productivity/workflows/teach/SKILL.md
productivity.to-questionnaire mixed productivity explicitly invoked discovery questionnaire for a known recipient, with explicit destination and write authority before artifact creation not_available realms/productivity/workflows/to-questionnaire/SKILL.md
productivity.wait-what read_only productivity explicitly invoked context-aware re-pitch of an explanation that did not land not_available realms/productivity/workflows/wait-what/SKILL.md
productivity.wayfinder mixed productivity durable destination, fog, decision, and frontier mapping for explicitly invoked multi-session efforts library/prompts/route-stubs/productivity-wayfinder.md realms/productivity/workflows/wayfinder/SKILL.md
productivity.write-a-skill mixed productivity create or revise a portable agent skill with progressive disclosure library/prompts/route-stubs/productivity-write-a-skill.md realms/productivity/workflows/write-a-skill/SKILL.md
productivity.writing-for-agents mixed productivity write or revise reliable documents that agents consume library/prompts/route-stubs/productivity-writing-for-agents.md realms/productivity/workflows/writing-for-agents/SKILL.md
transcription3 routes
transcription.meeting-transcript-decomposition read_only transcription bounded meeting transcript or meeting-note decomposition not_available realms/transcription/workflows/meeting-transcript-decomposition/SKILL.md
transcription.mono-mic-session-transcript read_only transcription bounded transcription or decomposition of an operator-provided mono mic session capture or transcript not_available realms/transcription/workflows/mono-mic-session-transcript/SKILL.md
transcription.pixel-recording-transcript read_only transcription bounded transcription or decomposition of an operator-provided Pixel recording or derived transcript not_available realms/transcription/workflows/pixel-recording-transcript/SKILL.md

Source Inputs

Input Path Status Used for
active_catalogue docs/reference/active-catalogue.md available catalogue_summary, unknown_status
ansa_scorecard docs/audits/ansa-scorecards/latest.yml available runs_and_pr_evidence, dashboard_readiness
outcome_learning docs/audits/outcome-learning/latest.yml available improvement_queue, chains_and_workbench_candidates
self_improvement_candidates docs/audits/self-improvement-candidates/latest.yml available improvement_queue
routes_registry library/registry/routes.yml available catalogue_summary, invocation_hints
prompts_registry library/registry/prompts.yml available invocation_hints
skills_registry library/registry/skills.yml available catalogue_summary
workflows_registry library/registry/workflows.yml available catalogue_summary
templates_registry library/registry/templates.yml available catalogue_summary
adapters_registry library/registry/adapters.yml available catalogue_summary
scripts_registry library/registry/scripts.yml available catalogue_summary
generated_skills_registry library/registry/generated-skills.yml available component_catalogue, invocation_hints
codex_staged_profile library/adapters/codex/staged-execution-profile.md available component_catalogue, agent_catalogue

Evidence Classification

Facts

  • Named source files were read with availability status recorded.
  • Registry-backed active component counts and route metadata are facts when registry files are available.
  • Scorecard sample, CI, review, validation, and runtime-proof counts are facts copied from scorecard output.

Derived

  • Dashboard readiness statuses are derived from #97 KPI wording plus existing scorecard/catalogue/learning evidence.
  • Receipt outcome-evidence coverage is derived from selected scorecard counts only.
  • Route-chain candidates are derived only when outcome-learning evidence already names repeated route chains.

Unknown

  • exact_cost
  • exact_total_tokens
  • last_used_per_component
  • last_used_per_component
  • limitations_when_not_declared_in_registry
  • maturity_for_non_declared_surfaces
  • maturity_for_non_pilot_surfaces
  • required_context_when_not_declared_in_registry
  • success_rate_per_component
  • usage_count_per_component
  • usage_count_per_component

Not Yet Measured

  • projection_consecutive_success_count
  • dashboard_metadata_coverage_percentage
  • token_burn_when_exact_observed_totals_absent
  • stale_or_unused_route_detection
  • operator_dashboard_action_success_rate