Skip to main content
Glama

DollhouseMCP

by DollhouseMCP
QA_FRAMEWORK_V1.6.0_COORDINATION.md6.64 kB
# QA Framework v1.6.0 Implementation Coordination **Date**: August 21, 2025 Evening **Branch**: `feature/qa-framework-v1.6.0-improvements` **Orchestrator**: Opus 4.1 **Objective**: Implement critical QA improvements for v1.6.0 release ## Mission Overview Implement the must-have and should-have QA improvements identified from PR #662 review to establish a solid QA process for the v1.6.0 release. ## Implementation Priority Order 1. **#663** - CI/CD Integration (develop branch only) 2. **#667** - Tool validation before testing 3. **#666** - Use centralized config file 4. **#669** - Complete deprecated tool removal 5. **#665** - Test data cleanup mechanism ## Agent Assignments ### 📋 Agent Registry | Agent ID | Task | Issue | Status | Files to Modify | |----------|------|-------|--------|-----------------| | CI-1 | Add QA to CI workflows | #663 | 🔴 Not Started | `.github/workflows/core-build-test.yml` | | TOOL-1 | Implement tool validation | #667 | 🔴 Not Started | `scripts/qa-*.js`, tool discovery | | CONFIG-1 | Wire up config file | #666 | 🔴 Not Started | `scripts/qa-*.js` to use `test-config.js` | | CLEANUP-1 | Remove deprecated tools | #669 | 🔴 Not Started | All QA scripts | | DATA-1 | Test data cleanup | #665 | 🔴 Not Started | `scripts/qa-github-integration-test.js` | Legend: 🔴 Not Started | 🟡 In Progress | 🟢 Complete | 🔵 Blocked ## Technical Context ### Current State - QA scripts exist but have issues: - Hardcoded timeouts instead of using config - No tool validation before testing - Some deprecated tool references remain - No cleanup of test data created - CI/CD doesn't run QA tests yet - `test-config.js` exists but isn't used ### CI/CD Strategy (Issue #663) Add to existing workflows with develop-only condition: ```yaml - name: Run QA Automation Tests if: github.base_ref == 'develop' || github.ref == 'refs/heads/develop' run: | node scripts/qa-direct-test.js node scripts/qa-github-integration-test.js continue-on-error: true # Non-blocking initially ``` ### Tool Validation Strategy (Issue #667) - Add tool discovery at start of each QA script - Filter out deprecated/non-existent tools - Only test tools that actually exist - Calculate success rates based on valid tools only ### Config Integration Strategy (Issue #666) - Import CONFIG from test-config.js - Replace all hardcoded timeouts (5000, 10000, 15000) - Use CONFIG.timeouts.tool_call, etc. ### Deprecated Tool Cleanup (Issue #669) Tools to remove references to: - `browse_marketplace` - `get_marketplace_persona` - `activate_persona` - `get_active_persona` - `deactivate_persona` ### Test Data Cleanup (Issue #665) - Track test personas/elements created - Add cleanup function at end of tests - Delete test artifacts after validation ## Agent Instructions ### For All Agents 1. **Scope Control**: Only modify files related to your specific task 2. **Test Your Changes**: Ensure changes don't break existing functionality 3. **Document Changes**: Add comments explaining what was changed and why 4. **Update This Document**: Mark your status when starting and completing ### Agent CI-1: CI/CD Integration **Objective**: Add QA tests to CI workflows for develop branch only **Tasks**: 1. Modify `.github/workflows/core-build-test.yml` 2. Add QA test step with develop-only condition 3. Make it non-blocking with `continue-on-error: true` 4. Test that it only runs on develop PRs **Success Criteria**: - QA tests run on develop branch PRs - Tests don't block merges (non-blocking) - Tests don't run on main branch ### Agent TOOL-1: Tool Validation **Objective**: Add tool discovery and validation to QA scripts **Tasks**: 1. Create a tool discovery function 2. Add to beginning of each QA script 3. Filter out non-existent tools before testing 4. Update success rate calculation **Success Criteria**: - Only existing tools are tested - Success rates reflect actual tool availability - Clear logging of skipped tools ### Agent CONFIG-1: Config Integration **Objective**: Wire up test-config.js to all QA scripts **Tasks**: 1. Import CONFIG from test-config.js in each script 2. Replace hardcoded timeout values 3. Use configuration constants throughout 4. Test that timeouts still work correctly **Success Criteria**: - No hardcoded timeouts remain - All scripts use centralized config - Tests still function properly ### Agent CLEANUP-1: Deprecated Tool Removal **Objective**: Remove all references to deprecated tools **Tasks**: 1. Search for deprecated tool names in all QA scripts 2. Remove or comment out deprecated tool tests 3. Update to use new element-based tools where applicable 4. Verify no deprecated tools remain **Success Criteria**: - No references to deprecated tools - Scripts only test current tools - Documentation updated ### Agent DATA-1: Test Data Cleanup **Objective**: Add cleanup for test artifacts **Tasks**: 1. Track all test data created (personas, elements) 2. Add cleanup function to delete test data 3. Ensure cleanup runs even on test failure 4. Test that cleanup works properly **Success Criteria**: - Test data is cleaned up after runs - No accumulation of test artifacts - Cleanup works even if tests fail ## Coordination Protocol ### Status Updates Agents should update their status in the registry when: - Starting work (🟡 In Progress) - Completing work (🟢 Complete) - Encountering blockers (🔵 Blocked) ### Communication - Document any dependencies between tasks - Note any issues discovered - Record what was changed and why ### Testing Each agent should: 1. Test their changes locally if possible 2. Ensure no regressions 3. Document test results ## Success Metrics ### Must Have for v1.6.0 - ✅ QA tests run automatically on develop PRs - ✅ Only valid tools are tested - ✅ Accurate success rate reporting - ✅ No test pollution between runs ### Quality Metrics - All hardcoded values removed - Clean, maintainable code - Clear documentation of changes - No breaking changes to existing functionality ## Risk Mitigation ### Potential Issues 1. **CI/CD changes break workflows** - Use continue-on-error 2. **Tool validation too strict** - Log skipped tools clearly 3. **Config changes break tests** - Test thoroughly 4. **Cleanup deletes wrong data** - Use specific test prefixes ### Rollback Plan - All changes on feature branch - Can revert individual commits if needed - Non-blocking CI changes won't affect main --- **Note**: This coordination document should be updated by each agent as they work. Opus will monitor progress but agents should work independently on their assigned tasks.

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/DollhouseMCP/DollhouseMCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server