- Removed 9 generated Istanbul coverage report files from
src/coverage/(HTML, CSS, PNG, JSON) - Added
.gitignorepatterns to prevent reappearance - None were imported by any TypeScript file
- Created
src/mcp/qe-tool-bridge.ts— adapter that registers non-overlapping QE tools - 11 tools were already exposed under flat names (e.g.,
test_generate_enhanced) - 37 total QE tools minus 11 skip = 26 tools registered via bridge
- These 26 tools were previously invisible to MCP clients despite being fully implemented
qe/tests/schedule— Test scheduling pipeline (phase scheduler, git-aware selector, flaky tracker)- Source:
src/test-scheduling/ - Calls
runTestPipeline()with realPipelineConfig
- Source:
qe/tests/load— Agent fleet load testing (light/medium/heavy profiles)- Source:
src/testing/load/ - Defaults to
mockMode=true(safe); user can setmockMode=falsefor real fleet testing
- Source:
qe/security/url-validate— URL security validation + PII exposure scanning- Checks for XSS, SQL injection, path traversal, open redirect patterns
- Scans URL/query params for PII (emails, SSNs, credit cards, phone numbers, API keys)
qe/workflows/browser-load— Browser workflow loader and validator- Loads from built-in YAML templates or inline YAML
- Validates workflow structure, interpolates variables into step configs
- Returns the resolved workflow ready for browser client execution
fleet_healthenrichment — Structural health via mincut-lambda- Passes actual agent nodes from
queen.getAgentsByDomain(), not an empty array - Falls back gracefully if native mincut module is unavailable
- Passes actual agent nodes from
- Deleted
src/cli/commands/qe-tools.ts(935 lines) —registerQEToolCommands()was never imported
- Moved
src/neural-optimizer/tosrc/_archived/neural-optimizer/ - Moved
tests/unit/neural-optimizer/totests/_archived/neural-optimizer/ - Excluded
src/_archived/from TypeScript compilation - Added
src/_archived/README.mdwith restore instructions
| Tool | Category |
|---|---|
fleet_init |
core |
fleet_status |
core |
fleet_health |
core (enriched with structural health) |
aqe_health |
core |
| Tool | Category |
|---|---|
task_submit |
task |
task_list |
task |
task_status |
task |
task_cancel |
task |
task_orchestrate |
task |
| Tool | Category |
|---|---|
agent_list |
agent |
agent_spawn |
agent |
agent_metrics |
agent |
agent_status |
agent |
team_list |
agent |
team_health |
agent |
team_message |
agent |
team_broadcast |
agent |
team_scale |
agent |
team_rebalance |
agent |
| Tool | Maps to QE Tool |
|---|---|
test_generate_enhanced |
qe/tests/generate |
test_execute_parallel |
qe/tests/execute |
coverage_analyze_sublinear |
qe/coverage/analyze |
quality_assess |
qe/quality/evaluate |
security_scan_comprehensive |
qe/security/scan |
contract_validate |
qe/contracts/validate |
accessibility_test |
qe/a11y/audit |
chaos_test |
qe/chaos/inject |
defect_predict |
qe/defects/predict |
requirements_validate |
qe/requirements/validate |
code_index |
qe/code/analyze |
| Tool | Domain |
|---|---|
qe/coverage/gaps |
coverage-analysis |
qe/requirements/quality-criteria |
requirements-validation |
qe/visual/compare |
visual-accessibility |
qe/learning/optimize |
learning-optimization |
qe/learning/dream |
learning-optimization |
qe/analysis/token_usage |
code-intelligence |
qe/planning/goap_plan |
coordination |
qe/planning/goap_execute |
coordination |
qe/planning/goap_status |
coordination |
qe/mincut/health |
coordination |
qe/mincut/analyze |
coordination |
qe/mincut/strengthen |
coordination |
qe/embeddings/generate |
learning-optimization |
qe/embeddings/compare |
learning-optimization |
qe/embeddings/search |
learning-optimization |
qe/embeddings/store |
learning-optimization |
qe/embeddings/stats |
learning-optimization |
qe/coherence/check |
quality-assessment |
qe/coherence/audit |
quality-assessment |
qe/coherence/consensus |
quality-assessment |
qe/coherence/collapse |
quality-assessment |
qe/qx/analyze |
quality-assessment |
qe/tests/schedule |
test-execution |
qe/tests/load |
test-execution |
qe/security/url-validate |
security-compliance |
qe/workflows/browser-load |
test-execution |
| Tool | Category |
|---|---|
memory_store |
memory |
memory_retrieve |
memory |
memory_query |
memory |
memory_delete |
memory |
memory_usage |
memory |
memory_share |
memory |
| Tool | Category |
|---|---|
model_route |
routing |
routing_metrics |
routing |
infra_healing_status |
infra-healing |
infra_healing_feed_output |
infra-healing |
infra_healing_recover |
infra-healing |
| Tool | Maturity | Notes |
|---|---|---|
qe/tests/schedule |
Full | Calls real runTestPipeline() — requires Vitest + git |
qe/tests/load |
Full | Mock mode by default; real mode requires fleet_init |
qe/security/url-validate |
Full | Regex-based URL security + PII scanning — no browser needed |
qe/workflows/browser-load |
Loader only | Loads/validates/resolves workflows — does not execute steps (no browser client) |
fleet_health structural |
Conditional | Requires @ruvector/mincut native module; silently skipped if unavailable |
- 14 governance integration files exist under
src/governance/ - Currently feature-flagged and not exposed as MCP tools
- Status: keep as-is until maturity documentation is available
src/domains/enterprise-integration/has coordination infrastructure- Future: expose dedicated MCP tools for enterprise workflow orchestration
qe/workflows/browser-loadcurrently only loads/validates — doesn't execute- Future: add
qe/workflows/browser-executethat requires a browser client (Vibium/Playwright)
- Archived at
src/_archived/neural-optimizer/ - Restore when 100+ agent fleets become common
- See
src/_archived/README.mdfor restore instructions