Change cleanup logic from "check and maybe kill" to "kill unconditionally".
The previous approach allowed agents to read stale output from lingering
external agent processes, causing context pollution and reuse of old data.
New behavior:
- Kill ALL background shells as the FIRST action
- Do NOT read output from stale shells (BashOutput forbidden)
- Do NOT check what processes are doing - just terminate them
- Only proceed to Step 1 after all shells are dead
Suppress the automatic "Review skipped" comments on PRs.
CodeRabbit can still be invoked on-demand with @coderabbitai review.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix: remove debug console.log statements from ui.js
* fix: add error handling and rollback for temp directory cleanup
* fix: use streaming for hash calculation to reduce memory usage
* refactor: hoist CustomHandler require to top of installer.js and ui.js
* fix: fail fast on malformed custom module YAML
User customizations must be valid - silent skip hides broken configs.
* refactor: use consistent return type in handleMissingCustomSources
* refactor: clone config at install() entry to prevent mutation
Adds support for delegating adversarial code reviews to external CLI agents
(Codex, Gemini, or Claude) when available. This provides independent, unbiased
code reviews from a different AI model.
Changes:
- Add invoke-bash and set-var tags to workflow.xml execution engine
- Add external_review_agents configuration to install-config.yaml
- Rewrite code-review workflow to detect and invoke external agents
- Cache agent detection in config.yaml to avoid repeated CLI checks
- Add fallback to built-in review if external agents unavailable/fail
- Update checklist to reflect new external agent workflow
External agent invocation:
- Codex: codex exec --full-auto "prompt"
- Gemini: gemini -p "prompt" --yolo
- Claude: claude -p "prompt" --dangerously-skip-permissions
Custom modules with module.yaml configuration prompts were not being
collected during installation. Added customModulePaths option to
ConfigCollector to resolve custom module paths from selectedFiles
and cachedModules sources.
- Add .coderabbit.yaml with minimal config and path instructions
- Exclude node_modules from review scope
- Document pilot research and conclusions in docs/planning/
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
- Skip module selection prompt during update/reinstall
- Keep all existing installed modules by default
- This prevents inquirer from showing modules as 'obsolete items' with confusing delete options
- Modules are now preserved during update/reinstall operations
- Fix getAgentsFromDir in bmad-artifacts.js to recursively scan subdirectories
- This ensures agents like cbt-coach and wellness-companion that are in subdirectories are properly found
- Agents now correctly get slash commands in .claude/commands/bmad/mwm/agents/
- All agents from the manifest now have corresponding IDE commands
- Updated getAgentsFromDir to search subdirectories for .md files
- Fixed installPath construction to include nested directory structure
- This ensures agents in subdirectories (like cbt-coach/cbt-coach.md) get added to agent-manifest.csv
- All agents now get proper CLI slash commands regardless of nesting depth
- Previously only scanned selectedModules, missing modules installed from custom locations
- Now scans the bmad directory to find all actually installed modules
- Any module with agents/workflows/tasks/tools will be included in manifests
- This fixes issue where MWM workflows weren't getting slash commands
- All modules now get equal treatment in IDE integration
sjennings
Alex Verkhovsky
Scott Jennings
Brian Madison
Murat K Ozcan
OhSeungWan
Wendy Smoak