@llipe.com/dev-tasks 0.8.0 → 0.9.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (99) hide show
  1. package/.claude/agents/developer.md +16 -11
  2. package/.claude/agents/qa-engineer.md +104 -0
  3. package/.claude/agents/verifier.md +1 -1
  4. package/.claude/commands/developer.md +14 -10
  5. package/.claude/commands/planner.md +10 -8
  6. package/.claude/commands/qa-engineer.md +23 -0
  7. package/.claude/skills/activity-coverage-gap-analysis/SKILL.md +107 -0
  8. package/.claude/skills/activity-test-implementation/SKILL.md +109 -0
  9. package/.claude/skills/activity-test-standards/SKILL.md +122 -0
  10. package/.claude/skills/implement/SKILL.md +14 -12
  11. package/.github/agents/developer.agent.md +16 -11
  12. package/.github/agents/planner.agent.md +10 -8
  13. package/.github/agents/qa-engineer.agent.md +104 -0
  14. package/.github/agents/verifier.agent.md +1 -1
  15. package/.github/instructions/implement.instructions.md +14 -12
  16. package/.github/prompts/qa-engineer.prompt.md +23 -0
  17. package/.github/skills/activity-coverage-gap-analysis/SKILL.md +107 -0
  18. package/.github/skills/activity-test-implementation/SKILL.md +109 -0
  19. package/.github/skills/activity-test-standards/SKILL.md +122 -0
  20. package/.kiro/agents/developer.md +16 -11
  21. package/.kiro/agents/planner.md +10 -8
  22. package/.kiro/agents/qa-engineer.md +108 -0
  23. package/.kiro/agents/verifier.md +1 -1
  24. package/.kiro/skills/activity-coverage-gap-analysis/SKILL.md +107 -0
  25. package/.kiro/skills/activity-test-implementation/SKILL.md +109 -0
  26. package/.kiro/skills/activity-test-standards/SKILL.md +122 -0
  27. package/.kiro/steering/implement.md +14 -12
  28. package/AGENTS.md +50 -23
  29. package/CHANGELOG.md +56 -0
  30. package/CLAUDE.md +6 -3
  31. package/README.md +108 -112
  32. package/TESTING.md +183 -0
  33. package/bundle-manifest.json +1 -0
  34. package/dist/core/distribution/install.d.ts.map +1 -1
  35. package/dist/core/distribution/install.js +25 -2
  36. package/dist/core/distribution/install.js.map +1 -1
  37. package/dist/core/distribution/profiles.d.ts +21 -0
  38. package/dist/core/distribution/profiles.d.ts.map +1 -1
  39. package/dist/core/distribution/profiles.js +21 -0
  40. package/dist/core/distribution/profiles.js.map +1 -1
  41. package/dist/core/distribution/update.d.ts +3 -0
  42. package/dist/core/distribution/update.d.ts.map +1 -1
  43. package/dist/core/distribution/update.js +141 -15
  44. package/dist/core/distribution/update.js.map +1 -1
  45. package/dist/core/extract/asyncapi/declared.d.ts +19 -0
  46. package/dist/core/extract/asyncapi/declared.d.ts.map +1 -0
  47. package/dist/core/extract/asyncapi/declared.js +168 -0
  48. package/dist/core/extract/asyncapi/declared.js.map +1 -0
  49. package/dist/core/extract/asyncapi/index.d.ts +1 -0
  50. package/dist/core/extract/asyncapi/index.d.ts.map +1 -1
  51. package/dist/core/extract/asyncapi/index.js +1 -0
  52. package/dist/core/extract/asyncapi/index.js.map +1 -1
  53. package/dist/core/extract/asyncapi/types.d.ts +1 -1
  54. package/dist/core/extract/asyncapi/types.d.ts.map +1 -1
  55. package/dist/core/extract/ladder.d.ts +79 -0
  56. package/dist/core/extract/ladder.d.ts.map +1 -0
  57. package/dist/core/extract/ladder.js +70 -0
  58. package/dist/core/extract/ladder.js.map +1 -0
  59. package/dist/core/extract/openapi/index.d.ts +35 -5
  60. package/dist/core/extract/openapi/index.d.ts.map +1 -1
  61. package/dist/core/extract/openapi/index.js +136 -3
  62. package/dist/core/extract/openapi/index.js.map +1 -1
  63. package/dist/core/extract/openapi/route2-introspect.d.ts +18 -0
  64. package/dist/core/extract/openapi/route2-introspect.d.ts.map +1 -0
  65. package/dist/core/extract/openapi/route2-introspect.js +249 -0
  66. package/dist/core/extract/openapi/route2-introspect.js.map +1 -0
  67. package/dist/core/extract/openapi/route2.d.ts +40 -17
  68. package/dist/core/extract/openapi/route2.d.ts.map +1 -1
  69. package/dist/core/extract/openapi/route2.js +202 -10
  70. package/dist/core/extract/openapi/route2.js.map +1 -1
  71. package/dist/core/extract/openapi/types.d.ts +3 -3
  72. package/dist/core/extract/openapi/types.d.ts.map +1 -1
  73. package/dist/core/extract/provider.d.ts +4 -0
  74. package/dist/core/extract/provider.d.ts.map +1 -1
  75. package/dist/core/extract/report.d.ts +29 -1
  76. package/dist/core/extract/report.d.ts.map +1 -1
  77. package/dist/core/extract/report.js +7 -2
  78. package/dist/core/extract/report.js.map +1 -1
  79. package/dist/core/extract/schema.d.ts +16 -18
  80. package/dist/core/extract/schema.d.ts.map +1 -1
  81. package/dist/core/extract/schema.js +48 -43
  82. package/dist/core/extract/schema.js.map +1 -1
  83. package/dist/core/extract/workspaces.d.ts +33 -0
  84. package/dist/core/extract/workspaces.d.ts.map +1 -0
  85. package/dist/core/extract/workspaces.js +193 -0
  86. package/dist/core/extract/workspaces.js.map +1 -0
  87. package/package.json +11 -2
  88. package/dist/core/extract/openapi/llm-descriptions.d.ts +0 -41
  89. package/dist/core/extract/openapi/llm-descriptions.d.ts.map +0 -1
  90. package/dist/core/extract/openapi/llm-descriptions.js +0 -36
  91. package/dist/core/extract/openapi/llm-descriptions.js.map +0 -1
  92. package/dist/core/extract/orm/llm-descriptions.d.ts +0 -28
  93. package/dist/core/extract/orm/llm-descriptions.d.ts.map +0 -1
  94. package/dist/core/extract/orm/llm-descriptions.js +0 -64
  95. package/dist/core/extract/orm/llm-descriptions.js.map +0 -1
  96. package/dist/core/extract/orm/migration-inference.d.ts +0 -25
  97. package/dist/core/extract/orm/migration-inference.d.ts.map +0 -1
  98. package/dist/core/extract/orm/migration-inference.js +0 -76
  99. package/dist/core/extract/orm/migration-inference.js.map +0 -1
@@ -0,0 +1,109 @@
1
+ ---
2
+ name: activity-test-implementation
3
+ description: "Author tests for Layer 1 and Layer 2 — unit, schema validation, backend component, mocked APIs, fixtures and gold datasets — with enforceable per-layer boundaries and a mandatory security-negative category. Use when writing or backfilling tests."
4
+ ---
5
+
6
+ # Activity: Test Implementation
7
+
8
+ Author tests for Layer 1 (deterministic foundations) and Layer 2 (constrained model and tool tests), with explicit boundaries for what belongs where. Invoked by the `qa-engineer` agent as step 2 of its procedure.
9
+
10
+ ---
11
+
12
+ > **RFC 2119 Notice:** The key words **MUST**, **MUST NOT**, **REQUIRED**, **SHALL**, **SHALL NOT**, **SHOULD**, **SHOULD NOT**, **RECOMMENDED**, **MAY**, and **OPTIONAL** in this document are to be interpreted as described in [RFC 2119](https://www.rfc-editor.org/rfc/rfc2119).
13
+
14
+ ## Goal
15
+
16
+ Produce tests that fail when behavior is wrong. Coverage is a by-product, not the objective: a suite can execute every line while asserting nothing that would catch a defect.
17
+
18
+ Consult `/TESTING.md` first. Its per-layer boundaries govern; this skill supplies the method.
19
+
20
+ ## Layer 1 — Deterministic Foundations
21
+
22
+ ### Unit tests
23
+
24
+ - **Scope:** one module's observable behavior. No I/O, no network, no real database, no clock dependence.
25
+ - **Isolation boundary:** stub at the module's own dependency edge, not three layers down. If a test needs four mocks to construct, the unit is too large or the boundary is wrong.
26
+ - **Test doubles:** a double stands in for a dependency. It **MUST NOT** reimplement the logic it replaces. A mock that grows a copy of production behavior means the test validates the mock.
27
+ - **Table-driven cases:** prefer one parameterized case list over many near-identical tests for input-domain variation.
28
+ - **Naming:** state the behavior and the condition — `rejects a negative page size`, not `test pagination 2`.
29
+ - **Assertions:** assert observable outcomes. Asserting the text of a generated query, or how many times an internal helper was called, couples the test to implementation without proving the behavior.
30
+
31
+ **Layer 1 MUST NOT:** open sockets or database connections, read the wall clock without injection, depend on test execution order, or assert on internal call counts as a proxy for behavior.
32
+
33
+ ### Schema validation
34
+
35
+ - Validate every externally-shaped payload against its schema: API request and response bodies, config files, event payloads, persisted documents.
36
+ - Required cases per schema: valid instance; missing required field; wrong type; unknown extra field (assert the documented policy, tolerate or reject); boundary values.
37
+ - Where a schema and a hand-written type can drift, assert they agree.
38
+
39
+ ## Layer 2 — Constrained Model and Tool Tests
40
+
41
+ ### Backend component tests
42
+
43
+ - **Scope:** a route handler, job, or service exercised through its real entry point, with the real in-process application wiring and a real datastore where feasible; only third-party externals are doubled.
44
+ - This layer exists because Layer 1 cannot see query syntax errors, wrong column types, `NULL` handling, or malformed joins. If the data layer is mocked everywhere, those defects are structurally invisible.
45
+ - Assert on the response or persisted state, not on the statement sent to the driver.
46
+
47
+ **Layer 2 MUST NOT:** reimplement in the test what the datastore does in production. Filtering, sorting, and pagination performed by a fake in the test language proves nothing about the query that ships.
48
+
49
+ ### Mocked APIs
50
+
51
+ - Double at the transport boundary (HTTP interception), not by replacing your own client module. Replacing the client skips serialization, headers, status handling, and error mapping.
52
+ - Required cases per external dependency: success; 4xx; 5xx; timeout; malformed body; empty body.
53
+ - Record which behaviors are asserted against a double rather than the real service, and treat that as a stated limitation.
54
+
55
+ ### Fixtures and gold datasets
56
+
57
+ - Shared fixtures and helpers live in one place per package. A token builder or client mock duplicated across files is a defect — consolidate it.
58
+ - Golden files: record how each was captured and how to regenerate it. State the source and date.
59
+ - Prefer asserting invariants and identities over long exact-value snapshots. Asserting `len(rows) == 84` against a real sample means replacing the sample breaks many tests at once for no behavioral reason.
60
+ - Derive fixture paths from a package-rooted anchor. Deriving them by walking up a fixed number of parent directories turns a moved package into a misleading "file not found".
61
+
62
+ ## Mandatory Security-Negative Category
63
+
64
+ **REQUIRED** for every authentication and authorization code path. Positive-path tests alone provide no security evidence.
65
+
66
+ | Case | Assertion |
67
+ | ---------------------------- | --------------------------------------------------------------- |
68
+ | **Invalid signature** | A token signed with the wrong key, or unsigned, is rejected. |
69
+ | **Expired credential** | A token past its expiry is rejected. |
70
+ | **Wrong issuer or audience** | A token issued for another party is rejected. |
71
+ | **Tampered claims** | A payload modified after signing is rejected. |
72
+ | **Cross-tenant access** | A caller cannot read or write another tenant's data. |
73
+ | **Missing credential** | An unauthenticated request is rejected. |
74
+ | **Insufficient permission** | An authenticated caller lacking the right is denied, not 404'd. |
75
+
76
+ ### The "tests faithful to insecure code" trap
77
+
78
+ A suite that asserts only the behavior the implementation happens to have will pass over a vulnerability. If the code merely decodes a token without verifying its signature, and the tests construct unsigned tokens, every test passes and nothing is verified.
79
+
80
+ You **MUST** treat this as a **finding**, not a pass. Report it as a security finding naming the module, the missing verification, and the tests that encode the permissive behavior. A green suite is not evidence.
81
+
82
+ Where isolation is asserted against test doubles rather than the real policy layer — application-level tenant checks standing in for database row-level security — you **MUST** record that the production policy is unverified.
83
+
84
+ ## Edge-Case Coverage per Behavior
85
+
86
+ For each behavior under test, cover: empty and missing input; boundary values on both sides; the maximum-size case; malformed encoding; and the documented failure mode. Where the code parses external documents, include corrupt, truncated, and access-protected inputs, and encoding variants beyond ASCII.
87
+
88
+ ## Effort Allocation
89
+
90
+ Weight test effort by risk and complexity, not by ease of writing. Thirty tests on pure helpers alongside four on the module handling multi-currency, multi-section parsing is an inversion. Report the imbalance when you find it.
91
+
92
+ ## Output
93
+
94
+ - Tests written or updated, by path and layer
95
+ - Layer-boundary violations found in existing tests
96
+ - Security-negative gaps, per auth path
97
+ - Behaviors asserted only against doubles, as stated limitations
98
+ - Duplicated fixtures or helpers recommended for consolidation
99
+ - Effort-allocation imbalances
100
+
101
+ ## Final Instructions
102
+
103
+ 1. You **MUST** consult `/TESTING.md` and respect its layer boundaries.
104
+ 2. You **MUST** write the security-negative cases for every auth path in scope.
105
+ 3. You **MUST** report a green suite over a permissive implementation as a finding.
106
+ 4. You **MUST NOT** author a double that reimplements the logic it replaces.
107
+ 5. You **MUST NOT** write placeholder assertions such as `expect(true)`.
108
+ 6. You **MUST** record which behaviors are verified only against doubles.
109
+ 7. You **MUST NOT** edit application source to make a test pass — report the needed change instead.
@@ -0,0 +1,122 @@
1
+ ---
2
+ name: activity-test-standards
3
+ description: "Establish and maintain a project's /TESTING.md contract, detect test-harness defects, and verify that every test package is reachable from the aggregate test command and the CI and deploy gates. Use when bootstrapping or refreshing the testing standard."
4
+ ---
5
+
6
+ # Activity: Test Standards
7
+
8
+ Establish or refresh `/TESTING.md` for a project, detect harness defects, and verify gate reachability. Invoked by the `qa-engineer` agent as step 1 of its procedure.
9
+
10
+ ---
11
+
12
+ > **RFC 2119 Notice:** The key words **MUST**, **MUST NOT**, **REQUIRED**, **SHALL**, **SHALL NOT**, **SHOULD**, **SHOULD NOT**, **RECOMMENDED**, **MAY**, and **OPTIONAL** in this document are to be interpreted as described in [RFC 2119](https://www.rfc-editor.org/rfc/rfc2119).
13
+
14
+ ## Goal
15
+
16
+ Produce a filled, accurate `/TESTING.md` for this project, plus a defect report. The contract is per-package: a repository mixing languages **MUST** be describable without forcing every package into one language's conventions.
17
+
18
+ ## Part 1 — Establish the standard for an existing project
19
+
20
+ Run every detection step. Report `not detected` explicitly rather than guessing.
21
+
22
+ ### 1. Inventory packages
23
+
24
+ Read the workspace definition (`pnpm-workspace.yaml`, `workspaces` in `package.json`, `pyproject.toml`, Cargo workspace, etc.). List every package. A single-package repository yields one entry.
25
+
26
+ ### 2. Detect framework and runner per package
27
+
28
+ Look for `vitest`, `jest`, `mocha`, `pytest`, `go test`, `cargo test`, and their config files. Record the runner per package, not per repository.
29
+
30
+ ### 3. Inventory scripts
31
+
32
+ Record the existing test, lint, format, typecheck, coverage, and audit commands per package, using whatever names the project actually uses. Compare against the canonical set in Part 3.
33
+
34
+ ### 4. Locate tests and naming patterns
35
+
36
+ Record test directory locations and file-naming conventions (`*.test.ts`, `test_*.py`, `*_test.go`). Note packages with source but no tests.
37
+
38
+ ### 5. Detect coverage tooling
39
+
40
+ Look for a coverage provider per package (`@vitest/coverage-v8`, `nyc`, `coverage.py`, `pytest-cov`). Record its absence explicitly — absence changes how the gap analysis runs, and it **MUST NOT** be silently inferred as zero coverage or as adequate coverage.
41
+
42
+ ### 6. Detect mocking and fixture approach
43
+
44
+ Record how test doubles are built, where fixtures live, and whether shared helpers exist. Flag duplicated helpers (the same token builder or client mock rewritten across files) as a defect.
45
+
46
+ ### 7. Fill `/TESTING.md`
47
+
48
+ Write detected values into the per-package table and the narrative slots.
49
+
50
+ - If `/TESTING.md` is absent, create it from the shipped placeholder.
51
+ - If it is present and filled, additions **MUST** be additive. You **MUST NOT** rewrite or delete consumer-authored content.
52
+ - If it is present but every slot is still an unfilled marker, report status `unfilled`. You **MUST** distinguish `unfilled` from `filled`: an empty contract is not an established standard, and downstream agents **MUST NOT** read it as permission.
53
+
54
+ ## Part 2 — Harness-defect detection
55
+
56
+ Evaluate every check. Report each finding with **file path and expected state** — never as a generic warning.
57
+
58
+ | Check | Defect condition |
59
+ | ---------------------------- | ------------------------------------------------------------------------------------------------------------------------------------- |
60
+ | **Test environment** | A package rendering DOM or browser components runs under a bare `node` environment with no `jsdom`/`happy-dom`. |
61
+ | **Test config presence** | A package has tests but no test config file, so it runs on defaults with no `setupFiles` and no global cleanup. |
62
+ | **Path-alias parity** | An alias is defined in `tsconfig.json` but absent from the test config. The first aliased import passes `tsc` and fails at test time. |
63
+ | **Global cleanup policy** | `restoreMocks` is not enabled, or a stubbed global (`fetch`, timers) is never restored. Fragile against pool changes. |
64
+ | **Runtime version parity** | Local, CI, and production runtimes differ. Record all three per package; divergence is a finding. |
65
+ | **Locale/timezone fixtures** | No locale or timezone fixture policy while the code formats dates, currency, or numbers. |
66
+ | **False-green placeholders** | Assertions such as `expect(true).toBe(true)` that report health without exercising anything. |
67
+
68
+ ## Part 3 — Script contract and gate reachability
69
+
70
+ A correct script **name** is not a working gate. Verify both.
71
+
72
+ ### Canonical names (JS/TS)
73
+
74
+ `lint`, `lint:fix`, `format`, `format:check`, `typecheck`, `test`, `test:unit`, `test:integration`, `test:e2e`, `test:coverage`, `audit`, `validate`. Report missing or misnamed scripts with the expected name.
75
+
76
+ Non-JS packages are **not** held to these names. Record their equivalent command per purpose instead.
77
+
78
+ ### Reachability checklist
79
+
80
+ - [ ] Every workspace package containing tests is reachable from the aggregate test command.
81
+ - [ ] No test package is silently omitted from the aggregate — an omission is a defect **even when every script name is canonically correct**.
82
+ - [ ] The CI workflow's test job invokes the aggregate command.
83
+ - [ ] The deploy workflow's quality gate invokes the aggregate command.
84
+ - [ ] A gate that is conditional, manually dispatched, or defaulted off is reported as **not automatically enforced**.
85
+ - [ ] No CI workflow present at all is reported as `no CI gate found` — an explicit finding, never an implicit pass.
86
+
87
+ Report per package: reached, or missed with the gate that fails to reach it.
88
+
89
+ ### Worked example
90
+
91
+ A monorepo defines `test:node` as `app && scraper && db` while the `api` package holds the largest suite. Every script name is canonical. The CI job and the deploy quality gate both run the aggregate.
92
+
93
+ Correct output: **defect** — `api` is unreachable from the aggregate; the CI gate and the deploy gate both let untested backend code through. Reporting "all script names conform" here is a false pass.
94
+
95
+ ## Failure Modes
96
+
97
+ | Condition | Required behavior |
98
+ | ----------------------------- | --------------------------------------------------------------------------- |
99
+ | No `package.json` or manifest | Report `not applicable` with reason. Do not crash, do not pass. |
100
+ | Malformed manifest | Report the parse failure with file and line. Do not report zero packages. |
101
+ | No CI workflow files | Report `no CI gate found`. |
102
+ | Very large workspace | Complete, or return a bounded partial result stating what was not analyzed. |
103
+ | Re-run with no changes | Produce no duplicate sections and no already-fixed defects. |
104
+
105
+ ## Output
106
+
107
+ - `/TESTING.md` status: `created` | `filled` | `unfilled` | `present`
108
+ - Per-package detection table
109
+ - Harness defects: each with file path and expected state
110
+ - Script contract: missing or misnamed scripts with expected names
111
+ - Reachability: packages reached, packages missed, CI gate status, deploy gate status
112
+ - What was not analyzed, and why
113
+
114
+ ## Final Instructions
115
+
116
+ 1. You **MUST** run every detection step and every harness check.
117
+ 2. You **MUST** report each defect with a file path and the expected state.
118
+ 3. You **MUST** verify reachability, not just script naming.
119
+ 4. You **MUST** preserve consumer-authored `/TESTING.md` content.
120
+ 5. You **MUST** distinguish an unfilled placeholder from a filled contract.
121
+ 6. You **MUST NOT** install dependencies or edit application source.
122
+ 7. You **MUST NOT** report a pass for anything you could not inspect.
@@ -43,6 +43,7 @@ This activity assumes:
43
43
  - If HEAD is already a valid feature branch (matching `issue/*` or `story/*`), proceed to step 3.
44
44
  - Branch format: `issue/<issue-number>-<short-description>` or `story/<id>-<short-description>`
45
45
  3. You **MUST** open a **draft Pull Request** by delegating to `github-ops` whenever possible.
46
+ - A PR requires at least one commit. Open it immediately after the first commit on the feature branch — never before — and you **MUST NOT** continue past that first commit without it.
46
47
  - Base branch is the default branch unless an orchestrating caller explicitly provides a base-branch override.
47
48
  - PR title **MUST** follow Conventional Commits (e.g., `feat: implement issue 37`).
48
49
  - PR description **MUST** include `Closes #<issue-number>`.
@@ -100,25 +101,26 @@ If `github-ops` delegation is unavailable in the current runtime, you **MUST** a
100
101
  - apply step executed only after user confirmation
101
102
  - post-apply verification recorded
102
103
  5. `technical-writer` validation **MUST** include a drift/stale-doc check report, and completion **MUST NOT** proceed while unresolved drift remains.
103
- 6. A `verifier` audit in `audit` mode **MUST** have run against the delivered implementation, with its human-readable summary posted to the issue/PR. This gate is mandatory and non-skippable — it **MUST NOT** be skipped regardless of drift findings, and drift findings reported by this audit **MUST NOT** block completion. This condition is satisfied once the audit has run and been posted; any resulting drift is handled by `product-engineer`'s `activity-drift-reconciliation` flow after this gate, not before it.
104
- 7. The PR **MUST** be converted from draft to ready for review.
105
- 8. The PR **MUST** be approved by the appropriate reviewer per the merge authority policy in `github-ops`:
104
+ 6. A `qa-engineer` pass **MUST** have run and `coverage_gate` **MUST** be recorded as `PASS`, `FAIL`, or `SKIPPED(<reason>)`. The reason **MUST** be non-empty when skipped; an omitted field is treated as incomplete. A `FAIL` or `SKIPPED` value does not block completion — only omission does.
105
+ 7. A `verifier` audit in `audit` mode **MUST** have run against the delivered implementation, with its human-readable summary posted to the issue/PR. This gate is mandatory and non-skippable — it **MUST NOT** be skipped regardless of drift findings, and drift findings reported by this audit **MUST NOT** block completion. This condition is satisfied once the audit has run and been posted; any resulting drift is handled by `product-engineer`'s `activity-drift-reconciliation` flow after this gate, not before it.
106
+ 8. The PR **MUST** be converted from draft to ready for review.
107
+ 9. The PR **MUST** be approved by the appropriate reviewer per the merge authority policy in `github-ops`:
106
108
  - PRs targeting an **integration branch**: `planner` reviews and approves.
107
109
  - PRs targeting **`main`**: the **user** reviews and approves.
108
- 9. The PR **MUST** be merged by the authorized party (planner for integration branches, user for `main`).
109
- 10. You **MUST NOT** close the GitHub Issue until the PR is approved **AND** merged.
110
- 11. You **MUST NOT** close the issue while the PR is still in draft or pending review.
111
- 12. You **MUST** notify the user when the PR is ready for review — explicitly inform them so they can review and merge.
110
+ 10. The PR **MUST** be merged by the authorized party (planner for integration branches, user for `main`).
111
+ 11. You **MUST NOT** close the GitHub Issue until the PR is approved **AND** merged.
112
+ 12. You **MUST NOT** close the issue while the PR is still in draft or pending review.
113
+ 13. You **MUST** notify the user when the PR is ready for review — explicitly inform them so they can review and merge.
112
114
 
113
115
  ---
114
116
 
115
117
  ## GitHub Execution Rules Summary
116
118
 
117
- | Phase | Rule |
118
- | ------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
119
- | **Before coding** | Confirm issue open → Create branch (`github-ops`) → Open draft PR (`github-ops`) → Sync checklists |
120
- | **During coding** | One sub-task at a time → Mark `[x]` locally + GitHub → Wait for approval |
121
- | **Before closing** | All ACs verified → Quality gates pass (test/lint/format/typecheck/audit) → migration confirmation/apply/verify (when applicable) → docs drift check clear → verifier audit run + summary posted (non-blocking on drift) → PR ready → Approved → Merged → Then close issue |
119
+ | Phase | Rule |
120
+ | ------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
121
+ | **Before coding** | Confirm issue open → Create branch (`github-ops`) → First commit → Open draft PR (`github-ops`) → Sync checklists |
122
+ | **During coding** | One sub-task at a time → Mark `[x]` locally + GitHub → Wait for approval |
123
+ | **Before closing** | All ACs verified → Quality gates pass (test/lint/format/typecheck/audit) → migration confirmation/apply/verify (when applicable) → docs drift check clear → coverage_gate recorded → verifier audit run + summary posted (non-blocking on drift) → PR ready → Approved → Merged → Then close issue |
122
124
 
123
125
  ---
124
126
 
package/AGENTS.md CHANGED
@@ -27,6 +27,19 @@ If `/DESIGN.md` is missing and the requested scope includes UI work, agents **MU
27
27
 
28
28
  ---
29
29
 
30
+ ## Testing Standard Contract (TESTING.md)
31
+
32
+ `/TESTING.md` is the canonical testing contract for a repository. It declares the layer taxonomy and what belongs at each layer, the per-package runner and commands, coverage thresholds and baseline policy, the fixture and mocking strategy, and the mandatory security-negative cases.
33
+
34
+ - `qa-engineer` **OWNS** `/TESTING.md` and fills it per project via `activity-test-standards`.
35
+ - `developer` **MUST** keep it current when the testing contract changes.
36
+ - Agents that read it **MUST** treat an unfilled placeholder as "no standard established", never as permission.
37
+ - The contract is **per-package**: a repository mixing languages (for example vitest and pytest) **MUST** be describable without forcing every package into one language's script names.
38
+
39
+ It ships with dev-tasks as a placeholder and is listed in `consumer_owned_paths`, so `dev-tasks update` never overwrites a filled version.
40
+
41
+ ---
42
+
30
43
  ## Taxonomy: Agent vs Skill vs Instruction
31
44
 
32
45
  | Concept | Purpose | Loaded When |
@@ -48,26 +61,32 @@ If `/DESIGN.md` is missing and the requested scope includes UI work, agents **MU
48
61
  | **housekeeping** | `housekeeping.agent.md` | Lint, type, and test-wiring fixes |
49
62
  | **github-ops** | `github-ops.agent.md` | GitHub consistency — standardizes issues, PRs, branches, labels, milestones, comments, and enforces merge authority policy |
50
63
  | **ux-engineer** | `ux-engineer.agent.md` | UX prototyping and gap analysis — turns PRD/SPEC into testable mockups and feeds refinements back to `product-engineer` |
64
+ | **qa-engineer** | `qa-engineer.agent.md` | Quality agent — establishes the `/TESTING.md` standard, authors tests for layers the project lacks, and reports coverage and structural gaps. Invoked by `developer` at the completion gate, before the `verifier` audit |
51
65
  | **verifier** | `verifier.agent.md` | Verification agent — owns compliance test-plan design (`design` mode) and post-implementation grey-box fidelity auditing (`audit` mode) against codebase, `/workstream`, tests, and PRD/spec intent |
52
66
 
67
+ `.github/agents/` and `.kiro/agents/` carry all nine agents. `.claude/agents/` carries seven by design: `planner` and `product-engineer` are orchestrators that must pause for user-approval gates, so on Claude Code they run in the main thread as `.claude/commands/` entry points rather than as subagents. See `CLAUDE.md`.
68
+
53
69
  ## Skills
54
70
 
55
71
  Skills are on-demand capabilities invoked by agents — **not** loaded unless explicitly referenced.
56
72
 
57
73
  ### Activity Skills
58
74
 
59
- | Skill | Directory | Purpose | Primary Consumer |
60
- | --------------------------------- | --------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- | ------------------ |
61
- | **activity-init** | `skills/activity-init/` | Establish product context and technical guidelines (mono/multi-repo/greenfield mode detection) | `product-engineer` |
62
- | **activity-refine** | `skills/activity-refine/` | Clarify scope — issue refinement or full PRD creation | `product-engineer` |
63
- | **activity-generate-spec** | `skills/activity-generate-spec/` | Transform PRD into technical specification | `product-engineer` |
64
- | **activity-generate-stories** | `skills/activity-generate-stories/` | Break spec into user stories with coverage validation | `product-engineer` |
65
- | **activity-publish-github** | `skills/activity-publish-github/` | Publish stories as GitHub Issues via MCP | `product-engineer` |
66
- | **activity-e2e-test-design** | `skills/activity-e2e-test-design/` | End-to-end black-box test scenario generation from spec/stories | `verifier` |
67
- | **activity-contract-test-design** | `skills/activity-contract-test-design/` | Consumer/provider contract and schema compatibility test strategy | `verifier` |
68
- | **activity-edge-case-refinement** | `skills/activity-edge-case-refinement/` | Systematic edge-case discovery by category with concrete examples | `verifier` |
69
- | **activity-random-test-tactics** | `skills/activity-random-test-tactics/` | Randomized, fuzz, and property-inspired test generation with reproducibility | `verifier` |
70
- | **activity-drift-reconciliation** | `skills/activity-drift-reconciliation/` | Routes verifier drift findings into task-list/checklist expansion, new issues, or PRD/spec changelog write-back (human-confirmed) | `product-engineer` |
75
+ | Skill | Directory | Purpose | Primary Consumer |
76
+ | ---------------------------------- | ---------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- | ------------------ |
77
+ | **activity-init** | `skills/activity-init/` | Establish product context and technical guidelines (mono/multi-repo/greenfield mode detection) | `product-engineer` |
78
+ | **activity-refine** | `skills/activity-refine/` | Clarify scope — issue refinement or full PRD creation | `product-engineer` |
79
+ | **activity-generate-spec** | `skills/activity-generate-spec/` | Transform PRD into technical specification | `product-engineer` |
80
+ | **activity-generate-stories** | `skills/activity-generate-stories/` | Break spec into user stories with coverage validation | `product-engineer` |
81
+ | **activity-publish-github** | `skills/activity-publish-github/` | Publish stories as GitHub Issues via MCP | `product-engineer` |
82
+ | **activity-e2e-test-design** | `skills/activity-e2e-test-design/` | End-to-end black-box test scenario generation from spec/stories | `verifier` |
83
+ | **activity-contract-test-design** | `skills/activity-contract-test-design/` | Consumer/provider contract and schema compatibility test strategy | `verifier` |
84
+ | **activity-edge-case-refinement** | `skills/activity-edge-case-refinement/` | Systematic edge-case discovery by category with concrete examples | `verifier` |
85
+ | **activity-random-test-tactics** | `skills/activity-random-test-tactics/` | Randomized, fuzz, and property-inspired test generation with reproducibility | `verifier` |
86
+ | **activity-test-standards** | `skills/activity-test-standards/` | Establish and maintain `/TESTING.md`, detect harness defects, verify test-script and CI/deploy gate reachability | `qa-engineer` |
87
+ | **activity-test-implementation** | `skills/activity-test-implementation/` | Author Layer 1-2 tests with enforceable boundaries and a mandatory security-negative category | `qa-engineer` |
88
+ | **activity-coverage-gap-analysis** | `skills/activity-coverage-gap-analysis/` | Measure coverage against a baseline, or run risk-ranked structural gap analysis when no provider exists | `qa-engineer` |
89
+ | **activity-drift-reconciliation** | `skills/activity-drift-reconciliation/` | Routes verifier drift findings into task-list/checklist expansion, new issues, or PRD/spec changelog write-back (human-confirmed) | `product-engineer` |
71
90
 
72
91
  ### Operational Skills
73
92
 
@@ -79,20 +98,27 @@ Skills are on-demand capabilities invoked by agents — **not** loaded unless ex
79
98
 
80
99
  ### Third-Party Skills
81
100
 
82
- | Skill | Directory | Purpose | Primary Consumer |
83
- | ------------------------------- | ------------------------------------- | ------------------------------------- | -------------------------- |
84
- | **vercel-composition-patterns** | `skills/vercel-composition-patterns/` | Vercel component composition patterns | `developer`, `ux-engineer` |
85
- | **vercel-react-best-practices** | `skills/vercel-react-best-practices/` | Vercel React best practices | `developer`, `ux-engineer` |
101
+ None currently vendored. Third-party skills, when added, are installed under each platform's skills directory alongside the skills above.
86
102
 
87
103
  ## Instructions (Scoped)
88
104
 
89
- Instructions are scoped via `applyTo`/`fileMatchPattern` and auto-applied to matching files.
105
+ Instructions are scoped via `applyTo`/`fileMatchPattern` and auto-applied to matching files. Claude Code has no scoped-instruction mechanism, so `plan` and `implement` ship there as skills instead.
106
+
107
+ | Instruction | Copilot / Kiro file | Claude equivalent | Scope | Purpose |
108
+ | --------------------------- | ------------------------------------------------------------------------------------------------------------------- | --------------------------- | -------------------------- | --------------------------------------------------------------------------------------------------------------- |
109
+ | **plan** | `.github/instructions/plan.instructions.md` / `.kiro/steering/plan.md` | `.claude/skills/plan/` | `workstream/**` | Convert stories or refined issues into execution-ready task lists |
110
+ | **implement** | `.github/instructions/implement.instructions.md` / `.kiro/steering/implement.md` | `.claude/skills/implement/` | `workstream/**/tasks-*.md` | Execute task list with step-gated approval, branching, and PR discipline |
111
+ | **nextjs-pages-components** | `.github/instructions/domain/nextjs-pages-components.instructions.md` / `.kiro/steering/nextjs-pages-components.md` | — | `**/app/**/*.tsx` | Next.js + React conventions |
112
+ | **git-guard-notice** | `.kiro/steering/git-guard-notice.md` | — | Always loaded (Kiro) | Restates the three git invariants: no push/merge to `main`, Conventional Commits, `--body-file` for `gh` bodies |
113
+
114
+ ## Hooks
115
+
116
+ | Hook | Files | Purpose |
117
+ | ---------------- | ---------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------- |
118
+ | **git-guard** | `.kiro/hooks/git-guard.json`, `.kiro/hooks/scripts/git-guard.sh`, `.claude/hooks/git-guard.sh` | Blocks pushes/merges to `main`, non-Conventional commit messages, and inline `gh --body` |
119
+ | **branch-guard** | `.kiro/hooks/scripts/branch-guard.sh` | Blocks write operations while on the default branch |
90
120
 
91
- | Instruction | File | Scope | Purpose |
92
- | --------------------------- | ------------------------------------------------ | -------------------------- | ------------------------------------------------------------------------ |
93
- | **plan** | `plan.instructions.md` | `workstream/**` | Convert stories or refined issues into execution-ready task lists |
94
- | **implement** | `implement.instructions.md` | `workstream/**/tasks-*.md` | Execute task list with step-gated approval, branching, and PR discipline |
95
- | **nextjs-pages-components** | `domain/nextjs-pages-components.instructions.md` | `**/app/**/*.tsx` | Next.js + React conventions |
121
+ Hook enforcement is best-effort. Human PR review is the actual gate for these invariants.
96
122
 
97
123
  ## Prompts
98
124
 
@@ -167,6 +193,7 @@ All AI coding agents working in this repository **MUST**:
167
193
  - Use canonical script names: `lint`, `format:check`, `typecheck`, `test`, `audit`, `validate`
168
194
  - Enforce quality gates before completion: `test`, `lint`, `format:check`, `typecheck`, `audit`
169
195
  - Use the `git-ops` skill for branch management, rebase, and conflict resolution
196
+ - `qa-engineer` runs at the completion gate, before the `verifier` audit, and records `coverage_gate`. It **MAY** be skipped only as `SKIPPED(<reason>)` with a non-empty reason; omitting the field is treated as incomplete.
170
197
  - The `verifier` audit is mandatory and non-skippable before every PR is marked ready. Drift findings are non-blocking and route to `product-engineer`'s `activity-drift-reconciliation`.
171
198
  - **Test-first design is the default:** `product-engineer` recommends `verifier` Design Mode after planning; `developer` writes tests before implementation code; `planner` checks for test plans and enforces test-first in developer handoffs
172
199
  - If `memo-cli` is installed and configured: agents **MUST** read/write entries per their role (see agent files for details)
@@ -266,4 +293,4 @@ When `scope.primary` contains exactly 1 component, no partitioning is needed. Th
266
293
 
267
294
  ---
268
295
 
269
- For workflow chains and sequencing diagrams, see [`docs/workflow-chains.md`](docs/workflow-chains.md).
296
+ For workflow chains and sequencing diagrams, see [`docs/workflow-chains.md`](docs/workflow-chains.md). For architecture, artifacts, and invariants, see [`docs/system-overview.md`](docs/system-overview.md) and [`docs/data-model.md`](docs/data-model.md). Full documentation index: [`docs/README.md`](docs/README.md).
package/CHANGELOG.md CHANGED
@@ -5,6 +5,61 @@ All notable changes to this project will be documented in this file.
5
5
  The format is based on [Keep a Changelog](https://keepachangelog.com/),
6
6
  and this project adheres to [Semantic Versioning](https://semver.org/).
7
7
 
8
+ ## [0.9.1] - 2026-08-18
9
+
10
+ ### Fixed
11
+
12
+ - fix(dist): discover new package files during update
13
+
14
+ ### Changed
15
+
16
+ - Merge pull request #134 from llipe:issue/fix-update-discover-new-files
17
+
18
+ ## [0.9.0] - 2026-08-18
19
+
20
+ ### Added
21
+
22
+ - feat(qa): wire TESTING.md distribution and update all registries
23
+ - feat(qa): wire qa-engineer into developer, planner, implement, verifier
24
+ - feat(qa): add three testing skills mirrored across platforms
25
+ - feat(qa): add TESTING.md contract and qa-engineer agent
26
+ - feat(extract): wire schema extraction through ladder (declared → observed)
27
+ - feat(extract): add pg as optional peer dep, update Known Limitations
28
+ - feat(extract): add AsyncAPI declared rung for on-disk spec detection
29
+ - feat(extract): add OpenAPI ladder orchestrator (declared → observed → inferred)
30
+ - feat(extract): remove in-CLI LLM, extend report with provenance and handoff
31
+ - feat(extract): add workspace discovery and ladder runner
32
+ - feat(extract): implement route2 spike — boot + introspect for Express
33
+
34
+ ### Fixed
35
+
36
+ - fix: format
37
+ - fix(qa): address Major audit findings D-1 and D-2
38
+ - fix(workflow): open draft PR after first commit, not before
39
+
40
+ ### Changed
41
+
42
+ - Merge pull request #129 from llipe:issue/123-qa-agent-and-testing-standard
43
+ - docs(qa): sync registries, guidelines, and ADR-003 for the coverage gate
44
+ - test(qa): add fixture projects reproducing the home-ledger defects
45
+ - test(dist): add failing root-file distribution assertions for TESTING.md
46
+ - test(qa): add failing testing-standard assertions and rule-19 guard
47
+ - test(qa): add failing qa-engineer parity and prompt-cap assertions
48
+ - Merge pull request #128 from llipe/docs/current-state-sync-post-125
49
+ - chore(workstream): reorganize docs into archive directory
50
+ - docs: sync documentation with current dt implementation state
51
+ - Merge pull request #125 from llipe:claude/dev-tasks-codebase-analysis-oxda22
52
+ - chore: resolve merge conflict with main in catalog index.yaml
53
+ - docs(workstream): fix manual test plan — scripts must run from project root
54
+ - docs(workstream): fix manual test plan — scripts must run from project root
55
+ - docs(workstream): rewrite manual test plan with proper tsx file approach
56
+ - docs(workstream): fix manual test plan for tsx top-level await limitation
57
+ - chore: fix lint and format issues, update task checklist
58
+ - chore(workstream): link task list to issue #127 and PR #125
59
+ - Merge pull request #126 from llipe:issue/docs-current-state-sync
60
+ - docs: reorganize and document system architecture and agent framework
61
+ - chore(workstream): add task list for extraction ladder inversion plan
62
+
8
63
  ## [0.8.0] - 2026-08-11
9
64
 
10
65
  ### Fixed
@@ -79,6 +134,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/).
79
134
  - feat(catalog): implement catalog query and routing commands (S-013) Implements dt catalog resolve, get, deps, consumers, flow, closure, and coverage subcommands with full --json support. - core/catalog/resolve.ts: text normalization (lowercase, de-accent, light es/en stemming, stopword removal) and weighted lexical scorer (exact id 100, alias 80, flow alias 75, domain 60, alias-token 40, name/desc 25) - core/catalog/queries.ts: graph reads (get, deps with depth/direction, consumers, flow, closure with include-consumers/max cap) - core/catalog/coverage.ts: extraction quality aggregation per component - adapters/cli/catalog-query.ts: CLI handler for all query subcommands - FlowEntry model extended with optional aliases field - Flow fixture files updated with aliases for routing tests Refs #46
80
135
  - feat(catalog): implement dt catalog validate with V01-V19 checks Implements referential integrity and structural validation checks per spec §6.2. Errors (V01-V11, V14-V15, V19) abort with exit 4; warnings (V12-V13, V16-V18) do not. Under --strict, V12 cycle warnings become errors. Supports allowed_cycles config. Closes #45
81
136
  - feat(catalog): support component.yaml in addition to component.json Build now looks for component.json first, then falls back to component.yaml. This matches the spec's mention of component.yaml while preserving backward compat with JSON format.
137
+ - **Correction (2026-08-14):** this fallback was reverted within the same release (`7bebc79` reverting `f697265`) and was never shipped. The released behavior of 0.7.0 is `component.json` only; a repo carrying only `component.yaml` is reported in `index.errors[]`. See [`docs/adr/ADR-001-component-json-manifest-format.md`](docs/adr/ADR-001-component-json-manifest-format.md).
82
138
  - feat(catalog): implement dt catalog build — aggregate manifests and generate index Implements Story S-011 (#43): - Define CatalogIndex types in core/catalog/index-model.ts - Implement registry parsing, manifest fetching, and mirroring - Build inverted consumer index from provides/consumes - Tally extraction quality from provenance fields - Aggregate domains and flows from manifests and catalog/flows/ - Generate deterministic catalog/index.yaml (sorted keys, stable order) - Implement idempotent write (skip when nothing changed) - Handle single repo failure: record in errors[], exit 3 - Detect duplicate component IDs across repos - Wire dt catalog build --registry <path> [--concurrency] [--json] CLI - Add 20-component fixture registry with diverse scenarios - Add yaml@2.7.1 dependency for YAML serialization Closes #43
83
139
  - feat(catalog): add JSON Schemas and dt validate-component Adds component.schema.json, flow.schema.json, and scope-output.schema.json (JSON Schema draft 2020-12) plus a local validate-component module using ajv's Ajv2020 class. Wires dt validate-component <path> with human and --json output, exit 0/4. No network access required. Refs #42
84
140
 
package/CLAUDE.md CHANGED
@@ -19,6 +19,8 @@ All work in this repository **MUST**:
19
19
  - Prefer `pnpm` over `npm` for JavaScript/TypeScript workflows (fallback to `npm` only when `pnpm` is unavailable or explicitly disallowed).
20
20
  - Use canonical `package.json` script names for JS/TS projects: `lint`, `lint:fix`, `format`, `format:check`, `typecheck`, `test`, `test:unit`, `test:integration`, `test:e2e`, `audit`, `validate`.
21
21
  - Before developer/planner completion, enforce and record quality gates: `test`, `lint`, `format:check`, `typecheck`, `audit`.
22
+ - `qa-engineer` runs at the completion gate, before the `verifier` audit, and records `coverage_gate: PASS | FAIL | SKIPPED(<reason>)`. Skipping requires a non-empty reason; omitting the field is treated as incomplete.
23
+ - `/TESTING.md` is the canonical testing contract — owned by `qa-engineer`, kept current by `developer`, and treated as "no standard established" while unfilled.
22
24
  - Reference GitHub Issues in branch names and commits; follow the `github-ops` conventions for all GitHub artifacts (issues, PRs, branches, labels, milestones, comments).
23
25
  - Treat `/DESIGN.md` as the source of truth for visual tokens, components, and design guidance; update it whenever UI contracts change.
24
26
  - Maintain document changelogs when updating generated artifacts (PRDs, specs, stories, refinements).
@@ -52,13 +54,14 @@ This toolkit is exposed as **commands** (entry points you invoke with `/`), **su
52
54
  | `/github-ops` | Audit/standardize GitHub artifacts. |
53
55
  | `/technical-writer` | Sync `/docs` with the codebase. |
54
56
  | `/housekeeping` | Fix lint/type/test-wiring issues. |
57
+ | `/qa-engineer` | Establish `/TESTING.md`, author missing Layer 1-2 tests, and report `coverage_gate` plus risk-ranked gaps. |
55
58
  | `/ux-engineer` | PRD/SPEC → React mockups + refinement handoff. |
56
59
  | `/verifier-design` | Generate a compliance test plan + traceability matrix from a spec/story (Design Mode). |
57
60
  | `/verifier-audit` | Grey-box fidelity audit of delivered work against requirements and PRD/spec intent (Audit Mode). |
58
61
 
59
62
  ### Subagents (`.claude/agents/`)
60
63
 
61
- `developer`, `github-ops`, `technical-writer`, `housekeeping`, `ux-engineer`, `verifier` — isolated, scoped-tool workers. Claude delegates to these automatically when a task matches their description, and the orchestrator commands (`/planner`, `/product-engineer`) drive them via the Task tool.
64
+ `developer`, `github-ops`, `technical-writer`, `housekeeping`, `ux-engineer`, `verifier`, `qa-engineer` — isolated, scoped-tool workers. Claude delegates to these automatically when a task matches their description, and the orchestrator commands (`/planner`, `/product-engineer`) drive them via the Task tool.
62
65
 
63
66
  ### Orchestration model (important)
64
67
 
@@ -69,9 +72,9 @@ Claude Code subagents **cannot spawn other subagents** — the hierarchy is flat
69
72
 
70
73
  ### Skills (`.claude/skills/`)
71
74
 
72
- Activity skills: `activity-init`, `activity-refine`, `activity-generate-spec`, `activity-generate-stories`, `activity-publish-github`, `activity-e2e-test-design`, `activity-contract-test-design`, `activity-edge-case-refinement`, `activity-random-test-tactics`, `activity-drift-reconciliation` (routes `verifier` drift findings into task-list/checklist expansion, new issues, or human-confirmed PRD/spec changelog updates — used by `product-engineer`). Process skills: `plan`, `implement` (the single source of truth for task-list execution rules, including the mandatory `verifier` audit gate). Operational: `git-ops`, `webapp-mockup`, `memo-cli-usage`.
75
+ Activity skills: `activity-init`, `activity-refine`, `activity-generate-spec`, `activity-generate-stories`, `activity-publish-github`, `activity-e2e-test-design`, `activity-contract-test-design`, `activity-edge-case-refinement`, `activity-random-test-tactics`, `activity-test-standards`, `activity-test-implementation`, `activity-coverage-gap-analysis`, `activity-drift-reconciliation` (routes `verifier` drift findings into task-list/checklist expansion, new issues, or human-confirmed PRD/spec changelog updates — used by `product-engineer`). Process skills: `plan`, `implement` (the single source of truth for task-list execution rules, including the mandatory `verifier` audit gate). Operational: `git-ops`, `webapp-mockup`, `memo-cli-usage`.
73
76
 
74
- `plan` and `implement` were the always-loaded Copilot instructions; as skills they load on demand instead of consuming context every turn.
77
+ `plan` and `implement` are scoped instructions on Copilot (`applyTo`) and scoped steering on Kiro (`fileMatchPattern`); as Claude skills they load on demand instead of consuming context on every matching turn.
75
78
 
76
79
  ---
77
80