agentic-qe 3.7.9 → 3.7.11

package/.claude/skills/wms-testing-patterns/scripts/validate-config.json

@@ -1,51 +1,51 @@
-{
-  "skillName": "wms-testing-patterns",
-  "skillVersion": "1.0.0",
-  "requiredTools": [
-    "jq"
-  ],
-  "optionalTools": [
-    "node",
-    "python3"
-  ],
-  "schemaPath": "schemas/output.json",
-  "requiredFields": [
-    "skillName",
-    "status",
-    "output",
-    "output.testingDomain",
-    "output.workflowCoverage"
-  ],
-  "requiredNonEmptyFields": [
-    "output.summary",
-    "output.testingDomain"
-  ],
-  "mustContainTerms": [
-    "warehouse",
-    "inventory"
-  ],
-  "mustNotContainTerms": [
-    "TODO",
-    "FIXME",
-    "placeholder",
-    "example.com",
-    "lorem ipsum"
-  ],
-  "enumValidations": {
-    ".status": [
-      "success",
-      "partial",
-      "failed",
-      "skipped"
-    ],
-    ".output.testingDomain": [
-      "inventory",
-      "pick-pack-ship",
-      "edi",
-      "rf-scanning",
-      "allocation",
-      "integration",
-      "mixed"
-    ]
-  }
-}
+{
+  "skillName": "wms-testing-patterns",
+  "skillVersion": "1.0.0",
+  "requiredTools": [
+    "jq"
+  ],
+  "optionalTools": [
+    "node",
+    "python3"
+  ],
+  "schemaPath": "schemas/output.json",
+  "requiredFields": [
+    "skillName",
+    "status",
+    "output",
+    "output.testingDomain",
+    "output.workflowCoverage"
+  ],
+  "requiredNonEmptyFields": [
+    "output.summary",
+    "output.testingDomain"
+  ],
+  "mustContainTerms": [
+    "warehouse",
+    "inventory"
+  ],
+  "mustNotContainTerms": [
+    "TODO",
+    "FIXME",
+    "placeholder",
+    "example.com",
+    "lorem ipsum"
+  ],
+  "enumValidations": {
+    ".status": [
+      "success",
+      "partial",
+      "failed",
+      "skipped"
+    ],
+    ".output.testingDomain": [
+      "inventory",
+      "pick-pack-ship",
+      "edi",
+      "rf-scanning",
+      "allocation",
+      "integration",
+      "mixed"
+    ]
+  }
+}

package/CHANGELOG.md

@@ -5,6 +5,42 @@ All notable changes to the Agentic QE project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [3.7.11] - 2026-03-06
+
+### Added
+
+- **Full @claude-flow/guidance governance integration** — All 8 governance modules now properly load AND wire their guidance counterparts using a local-first pattern where AQE logic is authoritative and guidance supplements:
+  - `continue-gate`: Maps `ContinueDecision` with full `StepContext` for budget-aware loop control
+  - `memory-write-gate`: Local-first contradiction detection, guidance supplements via `evaluateWrite()`
+  - `adversarial-defense`: `ThreatDetector.analyzeInput()` supplements local pattern detection; new `recordAgentInteraction()` and `detectCollusion()` methods via `CollusionDetector`
+  - `deterministic-gateway`: `gateway.evaluate()` supplements local tool-call allow/deny gating
+  - `proof-envelope`: Parallel audit trail mirroring via `ProofChain.append()`
+  - `shard-retriever`: Embedding-based relevance boosting via `ShardRetriever.retrieve()`
+  - `evolution-pipeline`: Signed change proposals via `EvolutionPipeline.propose()` on rule promote/demote
+  - `trust-accumulator`: Guidance trust scoring integration
+- **Governance subpath export** — Users can now `import {} from 'agentic-qe/governance'`
+
+### Fixed
+
+- **ContinueGate `mapGuidanceDecision` read wrong property names** — Was reading `decision.shouldContinue` / `.reason` / `.throttleMs` which don't exist on `ContinueDecision`; now correctly reads `decision.decision`, `decision.reasons`, `decision.metrics`
+- **MemoryWriteGate evaluation order** — Guidance was running before local logic; reversed to local-first so AQE's contradiction detection is authoritative
+- **ContinueGate `StepContext` was `Partial`** — Now provides full `StepContext` with realistic budget values derived from actual history
+- **Security: ReDoS guards** — Added line-length guards for all regex matches (CWE-1333)
+
+## [3.7.10] - 2026-03-05
+
+### Fixed
+
+- **MCP server entry path resolution** — `aqe mcp` command failed with "Could not find MCP server entry point" because `findMcpEntry()` traversed 2 directory levels instead of 1 from `dist/cli/` to find `dist/mcp/bundle.js`
+- **Stale v3/ directory references** — Worker daemon and MCP command lookup referenced `node_modules/agentic-qe/v3/dist/` which no longer exists after the v3.7.5 flatten; updated to `node_modules/agentic-qe/dist/`
+- **CRLF line endings breaking skill frontmatter** — 170+ skill files had Windows-style `\r\n` line endings causing `skill-lint` YAML parsers to fail on `---\r` delimiters; all converted to LF
+- **CI Node.js version** — Upgraded all 8 CI workflow files from Node.js 20 to Node.js 24
+
+### Changed
+
+- **README rewritten for clarity** — Reduced from 1,097 to ~280 lines; restructured around user outcomes instead of internal architecture details
+- **Asset build safety** — `prepare-assets.sh` now strips CRLF line endings and removes `.DS_Store` files as a safety net during asset preparation
+
 ## [3.7.9] - 2026-03-05
 
 ### Added