Skip to main content

Repo Gap Analysis

The implementation backlog for moving the repo closer to the ownerless governance standard. Every surface should be discoverable, validated, repairable, bounded, and purgeable. Audit dates: prior baseline 2026-05-14; refresh 2026-05-23 after 10 file-level slice audits; consolidation pass 2026-05-25 shipped the items below.

Resolved 2026-05-25 (consolidation pass)

This session closed the following items off the backlog:
  • P0 cron-is-dry-run bug – fixed in commit e42946cdf on docs-v2-dev-draft (inverted FLAGS default in all 6 dispatch-{concern}.yml scheduled jobs; manual-repair jobs unchanged). Pending merge to docs-v2. Unblocks 6+ stale data pipelines (contracts addresses, llms.txt, sitemap-ai.xml, OG/SEO generators).
  • P1 false-flag: decisions-log lags implementation – verified all 18 decisions present, split across 2 files: decisions-log.mdx holds D-ACT-01..10 + D-GOV-07..08; design/governance/design-overview.md holds D-GOV-01..06 (locked 2026-04-04). Was a paginated-read artefact, not a real gap.
  • P1 false-flag: .github/scripts/ empty – reframed: the folder is intentionally empty per locked decision D-ACT-06 (every script migrated to operations/scripts/ for governance reach). Downgraded to P2 path-drift cleanup for stale references in older docs.
  • P1 ownerless framing refresh post-D-ACT-10 – corrected “8 of 28 surfaces” claim across 6 canonical pages to live state: 4 of 5 unified surfaces ownerless-ready in repo-governance-surfaces.json (canonical post-D-ACT-10; legacy 8-entry ownerless-governance-surfaces.json superseded but kept on disk for reference).
  • P3 frameworks/github-actions.mdx v2.1 → v2.2 – synced to the 2026-05-22 4-tier composable refactor + D-ACT-09/10 + D-GOV-07/08; bumped lastVerified to 2026-05-25; refreshed Tier-table counts (6+5=11 workflows, 230 atomics); added paragraph reconciling the “6 actions” framework framing with automations.mdx’s “11 active” framing.
  • P4 documentation gap: new dispatcher tranches – documented the three new veracity dispatchers (docs-research-packet.js, orchestrator-guides-research-review.js, docs-page-research-pr-report.js) under a new “Recently shipped (May 2026)” section in automations.mdx; also documented the local-only social-feed dispatcher (run-solutions-social-fetch.js) in contributor-tools.mdx.
  • Frontmatter compliance sweep on 33 canonical pages – brought all 16 frameworks + all 17 policies to canonical frontmatter (pageType / audience / purpose / status / lastVerified). All 33 now lint-clean (0 blocking, 0 warnings). Migrated 8 pageType: overview instances to canonical enum values across features pages + catalog.
The items below remain open. The cron P0 fix above means the next cron after merge to docs-v2 should also resolve the Stale generated governance map P0 and most data-staleness P0/P1 entries – verify after merge. Audit dates: prior baseline 2026-05-14; refresh 2026-05-23 after 10 file-level slice audits across docs-guide, ai-tools, operations/governance, .github, agent adapters, snippets, v2/internal + per-tab workspaces, per-tab checks.mdx cross-tab comparison, the 14 workspace plans (ORCHS, TERMINOLOGY, CONTRACTS, TOOLING, etc.), workspace history + BACKLOG, and root files. Evidence sources (live commands, 2026-05-23):
Detailed evidence in workspace/thread-outputs/repo-consolidation-deep/SLICE-*.md files (10 slices, ~4,400 lines of file-level inventory).

Executive findings – what’s P0 and P1

Purge + remediation queue

1. Governance artefacts

2. V2 folder governance

Cleanup matrix totals unchanged from 2026-05-14: Implementation order:
  1. Update .mintignore + shared selection helpers so _workspace, x-deprecated, archive, and known transitional lanes are consistently non-publishable.
  2. No-write report; split rows into move waves by section.
  3. Start with gateways (199 rows; carries most x-resources and x-deprecated lanes).
  4. Use git mv or repo move tooling – never hand-copy.
  5. Preserve route behaviour: if a page was ever routable, verify redirects or docs.json removal explicitly.
  6. Regenerate page/catalog indexes after each wave.
  7. lpd test --staged + targeted link/import validation.
Acceptance: no publishable route resolves into _workspace, x-deprecated, archive, or design lanes; file-selection helpers share the same non-publishable lane contract; matrix rows decrease after each wave.

3. Reports, archives, backup artefacts

Script footprint audit: 4 medium + 20 low (no critical/high). Specific purge candidates:

4. Scripts and tooling

5. GitHub Actions

6. Components, templates, UI

7. Per-tab content + IA

8. v2/internal + RFP

9. AI features + skills

10. Workspace plans

11. Tab gates stuck at Phase 2

Per workspace/plan/active/CONTENT-WRITING/decisions/tab-status.md:
  • 0 of 5 tabs has IA Approved – all ⬜
  • 3 of 5 tabs have Content Scan Done ✅ (Gateways, About, Delegators)
  • 0 of 5 tabs have advanced past Phase 2
The project pipeline is stalled. A tab is only “open” for the next phase when its gate is explicitly unlocked by human sign-off recorded in decision-registry.md. Action: human IA review + lock per tab to unblock downstream phases.

Implementation plan (refreshed 2026-05-23)

Phase 1: Stop the bleed (P0)

  1. Fix the cron-is-dry-run bug. Single edit across 6 dispatch-{concern}.yml files. Unblocks contracts, llms.txt, sitemap-ai.xml, OG-images, SEO, glossary regeneration.
  2. Regenerate stale governance map. Run the generator + commit.
  3. Wire OpenAPI fetcher to all 5 specs + schedule + validation gate.
  4. Fix docs.json default version (v1 → v2 OR confirm Mintlify dashboard override).
  5. Refresh stale RFP/internal claims – sibling pages cite 3 different script counts.
  6. Regenerate TAB-SUMMARY.md files to reflect populated internal stubs.
Exit criteria: governance map --check passes; data-pipeline staleness drops to ≤1 day; AI artefacts regenerate on every docs.json change; visitors land on v2 by default.

Phase 2: Repair the contracts (P1)

  1. Add /compliance/ to agent-governance-framework.mdx validator path references.
  2. Update contracts pipeline doc to reference current script location.
  3. Wire CI cross-index check across the 3 decision registries.
  4. Append D-ACT-05..10 + D-GOV-01..08 entries to decisions-log.mdx.
  5. Update framework-canonical.md published version to match 2026-05-22 refactor.
  6. Verify D-ACT-03 social-data consolidation completion (target: update-data-feeds.yml exists or scripts under dispatch/copy/).
Exit criteria: zero stale path references; published frameworks match live state; CI catches future drift.

Phase 3: Normalise non-publishable lanes (P1)

  1. Update .mintignore + selection helpers.
  2. Execute v2 cleanup matrix by section, starting with gateways (199 of 307 rows).
  3. Validate route + link + import behaviour after each wave.
  4. Record waves in a report under workspace/reports/repo-ops/.
Exit criteria: cleanup matrix decreases; no routed page in non-publishable lanes; archive naming normalised to x-archive/.

Phase 4: Resolve per-tab canonical contradictions (P1)

  1. Decide gateways canonical: produce gateways-specific OR formally adopt orchestrators’s as shared (declared).
  2. Add Cat 10 to gateways checks.mdx.
  3. Add frontmatter + run em-dash remediator on orchestrators + gateways checks.mdx.
  4. Lock bash-block policy site-wide.
  5. Lock persona enumeration in Frameworks.mdx §1.3.
  6. Promote Frameworks.mdx §1.6 InfoType from Draft.
  7. Move/symlink ORCHS tab map to canonical location; produce for other 4 tabs.
Exit criteria: per-tab canonical artefacts agree on shared spec; no checks.mdx violates its own Cat 1.1 or 2.12; Phase 0 pre-flight passes for at least 1 tab.

Phase 5: Component + skill backlog (P1)

  1. Replace hand-authored component count tables with generator-emitted blocks (4 surfaces).
  2. Fix snippets-registry.mdx lines 868-878 wrapper/displays label bug.
  3. Move 6 wrapper-tagged components OR update their @type JSDoc.
  4. Triage 24 archived JSX components (DELETE or KEEP-AS-ALIAS).
  5. Move all 8 Tier-1 composables out of composables/pages/unclassified/.
  6. Move 4 data files out of composables/.
  7. Delete 4 byte-identical template duplicate pairs.
  8. Promote 1 draft SKILL.md per week through 3-pass validation (18 in queue).
  9. Reconcile portable vs local skill divergence (34 vs 53; 38 portable-only with no local).
  10. Trim .claude/CLAUDE.md to ≤200 lines.
Exit criteria: every count generator-sourced; 0 wrapper-tagged components in displays/; 0 draft skills in production use; thin-adapter compliance per D-DG-11.

Phase 6: Backfill + retention (P1/P2)

  1. Backfill the 45 retired-@category + 4 retired-@domain scripts.
  2. Update .githooks/pre-commit + pre-push JSDoc to canonical 11-tag.
  3. Wire .githooks/verify.sh back into pre-commit OR retire.
  4. Apply report retention policy.
  5. Delete 13-file cleanup queue (.DS_Store, .bak, .legacy-duplicate.json, etc.).
  6. Untrack 1,070 archived locale snapshots + 7+ committed .zip files.
  7. Delete 20 byte-identical audit alias files.
  8. Implement tasks-retention.yml workflow for _workspace/ TTL enforcement.
Exit criteria: script footprint audit @category + @domain counts at 0; hook scripts pass the validator they enforce; workspace report bloat reduced; retention TTL enforced.

Non-acceptance criteria

The following do not satisfy this gap analysis:
  • Adding ignore rules without moving, validating, or documenting the underlying source of drift
  • Deleting files without dependency checks and required deletion approval
  • Hand-editing generated outputs without fixing the generator or canonical source
  • Leaving old docs paths as undocumented compatibility aliases
  • Treating workspace reports as permanent source of truth
  • Running only a superficial command when repo tooling provides a deeper validator
  • Marking a fix done without verifying the prior failure mode no longer reproduces

Next concrete commands

  • Adaptive Architecture – the model this gap analysis remediates against
  • Automations – the cron-dry-run bug that gates Phase 1
  • Feature Map – the canonical inventory cross-link
  • Ownerless Governance Policy – the 4-part contract that defines “complete”
  • Workspace Lifecycle Policy – TTL rules + tasks-retention.yml target
  • Cleanup + Quarantine Policy – deletion-discipline contract
  • workspace/plan/future/BACKLOG/master-tasks.md – BL-001..BL-046 with priorities + status + dates
  • workspace/thread-outputs/repo-consolidation-deep/ – 10 file-level slice audits (~4,400 lines of evidence)
  • workspace/thread-outputs/repo-consolidation-deep/SLICE-00-MASTER.md – master consolidation map (input to this gap analysis)
Last modified on June 2, 2026