Release-quality evaluation harness and rollout gate for Claude Skills
Last Updated: 2026-07-10
| # | File | Description |
|---|---|---|
| 001 | 001-PP-BCASE-business-case.md | Business case |
| 002 | 002-PP-PRD-product-requirements.md | Product requirements |
| 004 | 004-PP-UJRN-user-journey.md | User journey |
| 007 | 007-PP-PLAN-master-build-blueprint.md | Master build blueprint (canonical) |
| # | File | Description |
|---|---|---|
| 003 | 003-AT-ARCH-architecture.md | Architecture |
| 005 | 005-AT-SPEC-technical-spec.md | Technical spec |
| 010 | 010-AT-SPEC-eval-spec-and-contract-guide.md | Eval spec + contract authoring guide |
| 018 | 018-AT-SPEC-pb7-adapter-measurement-protocol-2026-05-12.md | PB-7 provider-adapter measurement protocol (gating doc; ISEDC v1 Q5 CTO binding) |
| 021 | 021-AT-ARCH-repo-blueprint-2026-06-10.md | Per-repo blueprint (Blueprint-C application) |
| 023 | 023-AT-SPEC-ciso-gate-failure-modes-2026-06-15.md | CISO PASS/FAIL gate failure-mode reference |
| 026 | 026-AT-SPEC-refiner-core-api-2026-06-20.md | @intentsolutions/refiner-core API spec (value types, pure fns, RefinerStrategy) |
| 029 | 029-AT-SPEC-refiner-core-api-2026-07-08.md | @intentsolutions/refiner-core public API spec (documents the v0.2.0 surface) |
| # | File | Description |
|---|---|---|
| 006 | 006-OD-STAT-status.md | Status |
| 008 | 008-OD-REPT-release-v0.2.7.md | Release report — v0.2.7 |
| 009 | 009-OD-REPT-epic-01-aar.md | Epic 01 after-action report |
| 011 | 011-OD-REPT-epic-02-aar.md | Epic 02 after-action report |
| 012 | 012-OD-REPT-epic-03-aar.md | Epic 03 after-action report |
| 013 | 013-OD-REPT-epic-04-aar.md | Epic 04 after-action report |
| 014 | 014-OD-REPT-epic-05-aar.md | Epic 05 after-action report |
| 015 | 015-OD-REPT-epic-06-aar.md | Epic 06 after-action report |
| 016 | 016-OD-REPT-epic-07-aar.md | Epic 07 after-action report |
| 017 | 017-OD-REPT-epic-08-aar.md | Epic 08 after-action report |
| # | File | Description |
|---|---|---|
| 010 | 010-TQ-SOPS-audit-harness-baseline-2026-05-01.md | Audit-harness testing baseline (Intent Solutions Testing SOP) |
Numbering note: two documents share sequence
010(010-AT-SPEC-…and010-TQ-SOPS-…) — a filing collision predating this index. Both are retained as-is; renumbering would break inbound references. Next new doc uses030.
| # | File | Description |
|---|---|---|
| 019 | 019-AA-AUDT-appaudit-devops-playbook.md | Operator-grade DevOps playbook (appaudit) |
| 020 | 020-AA-AACR-release-hardening-iep-P2-2026-05-21.md | IEP Priority 2 release-hardening AAR |
| 022 | 022-AA-AACR-staging-stays-staging-forensics-2026-06-11.md | Staging-stays-staging production-Rekor promotion forensics |
| 024 | 024-AA-AACR-real-provider-dogfood-2026-06-17.md | Real-provider behavioral dogfood AAR (iaj-E10) |
| 025 | 025-AA-AACR-configurable-openai-compatible-provider-2026-06-16.md | Configurable OpenAI-compatible provider AAR (DeepSeek / Kimi / OpenRouter) |
| 027 | 027-AA-AACR-refiner-v0.1.0-release-2026-06-21.md | Skill Refiner npm release — @intentsolutions/refiner-core + @intentsolutions/refiner v0.1.0 (SLSA provenance) |
| # | File | Description |
|---|---|---|
| 028 | 028-DR-FIND-jrig-eval-criteria-ids-and-deterministic-validation-2026-06-28.md | Deep-dive: two j-rig eval-engine bugs (criteria-id binding + deterministic validation) that inflated evals into false NO-SHIPs |
Durable epic reference files in epics/:
| Epic | File |
|---|---|
| Epic Index | Table of contents and active epic |
| Epic 01 | Repo Foundation and Operating Standard |
| Epic 02 | Spec Layer and Contract System |
| Epic 03 | Package Integrity and Deterministic Checks |
| Epic 04 | Evidence Layer, Persistence and Run Lifecycle |
| Epic 05 | Trigger Harness and Skill Roster Simulation |
| Epic 06 | Functional Execution Harness and Observation Layer |
| Epic 07 | Judgment Layer, Calibration and Model Matrix |
| Epic 08 | Regression Baseline, Scoring CLI and CI Gate |
| Epic 09 | Optimizer and Experiment Engine |
| Epic 10 | Team/Product Eval Packs and Drift Operations |
Local, self-contained library of templates, reference standards, agent patterns, and epic workflow diagrams.
Templates (templates/)
| Directory | Contents | Source Tier |
|---|---|---|
skill-templates/ |
6 SKILL.md structural patterns | Tier 1 (skill-creator) |
eval-schemas/ |
Eval JSON schemas (evals.json, etc.) | Tier 1 (skill-creator) |
References (references/)
| Directory | Contents | Source Tier |
|---|---|---|
skill-standards/ |
AgentSkills.io spec, source-of-truth, frontmatter, validation | Tier 0–1 |
eval-patterns/ |
Eval methodology, workflows, output patterns | Tier 1 |
agents/ |
Grader, comparator, analyzer agent patterns | Tier 1 |
enterprise-standards/ |
100-point rubric, production validator schema registry | Tier 0, 3 |
drift-and-consistency/ |
Drift categories, source-of-truth hierarchy | Tier 1 |
epic-workflows/ |
10 ASCII workflow diagrams | Tier 3 (authored) |
- Total documents: 30 numbered docs (
001–029, with two010s) + 10 epics + templates & references library - Categories used: 6 — PP, AT, OD, TQ, AA, DR
- Next sequence number: 030
- Note: every per-doc file in this directory is the canonical source; this index is the navigation layer. Rebuild via
/doc-filing(or/validate-consistency, which flags index drift) when docs are added.