**Problem:**
Agent compiler was stripping TTS data (intro and voices) when converting
.agent.yaml files to .md files. This caused all agents to use fallback
voice (en_US-lessac-medium) and generic intros in party mode.
**Root Cause:**
buildFrontmatter() only included name and description fields, omitting
the tts section from source YAML.
**Fix:**
- Modified buildFrontmatter() to accept and include TTS data
- Updated compileToXml() to extract TTS from parsed YAML
- TTS data now preserved in compiled .md frontmatter
**Impact:**
- Each agent now gets unique voice (ryan-high, kristin-medium, etc.)
- Personalized intros ("Hi! I'm John..." instead of "Hi! I'm Pm...")
- Party mode agents now distinguishable by voice
**Testing:**
✅ Tested with pm.agent.yaml → correct voice: en_US-ryan-high
✅ Tested with analyst.agent.yaml → correct voice: en_US-kristin-medium
✅ Verified TTS frontmatter structure matches manifest-generator expectations
**Related:**
- Fixes voice assignment issue in PR #987
- Enables proper multi-agent party mode experience
- Manifest generator already handles TTS correctly (no changes needed)
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
CRITICAL BUG FIX: Party mode was showing agent identifiers like
"Analyst", "Architect", "Dev" instead of actual persona names like
"Mary", "Winston", "Amelia" due to incorrect field references.
Changes:
- step-01-agent-loading.md: Use [displayName] not [Agent Name]
- step-02-discussion-orchestration.md: Use [displayName] not [Agent Name]
- Added CRITICAL warnings to use displayName field from manifest CSV
This ensures agents appear with their proper persona names throughout
party mode conversations and TTS announcements.
Resolves AgentVibes test31/test32 user feedback where all agents
showed generic title-cased identifiers instead of personas.
Automatically includes --with-audio parameter when installing AgentVibes
for the first time (detects by checking if .agentvibes directory exists).
This ensures new users get audio tracks downloaded automatically.
Updates skip --with-audio to avoid re-downloading existing audio files.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Addresses code review feedback - removing (Optional) prevents agents from
skipping the section without evaluating the conditional logic within.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- 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>
- Remove 'elevenlabs' fallback from providerVoiceField mapping
- ElevenLabs is no longer supported in AgentVibes
- Only Piper and macOS providers remain
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Add optional `tts` section to agent schema with intro and voices
- Update manifest-generator.js to read TTS from YAML and generate provider-aware CSV
- Add TTS configuration to all 9 agent YAML files with Piper and macOS voices
- Voice map CSV now auto-generated from agent YAML instead of hardcoded
- Provider-aware: auto-selects correct voice based on active TTS (Piper/macOS)
Benefits:
- Declarative: voice config lives with agent definition
- Loose coupling: AgentVibes reads CSV, no YAML dependency
- Extensible: new agents automatically get voice mapping
Related:
- Created comprehensive test suite in AgentVibes repo (test/unit/bmad-voice-map.bats)
- All 10 tests passing (8 pass, 2 skip future features)
- Validates BMAD → AgentVibes voice configuration pipeline
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 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
- Added logic to create customize template files during agent compilation
- ModuleManager was only using existing customize files, not creating them
- Now customize.yaml files will be created for all module agents
- This fixes issue where agents in subdirectories had no customization support
Next: Need to fix agent-manifest.csv to find agents in subdirectories
- Module agents now discovered recursively at any depth in agents folder
- .agent.yaml files are compiled to .md format during module installation
- Custom agents also support subdirectory structure
- Agents maintain their directory structure when installed
- YAML files are skipped during file copying as they're compiled separately
- Added compileModuleAgents method to handle YAML-to-MD compilation
- Updated discoverAgents to recursively search for .agent.yaml files
- Agents in subdirectories are properly placed in _cfg/agents with relative paths
This fixes issue where agents like cbt-coach were not being compiled
and were only copied as YAML files.
- Module discovery now scans entire project recursively for install-config.yaml
- Removed hardcoded module locations (bmad-custom-src, etc.)
- Modules can exist anywhere with _module-installer/install-config.yaml
- All modules treated equally regardless of location
- No special UI handling for 'custom' modules
- Core module excluded from selection list (always installed first)
- Only install-config.yaml is valid (removed support for legacy config.yaml)
Modules are now discovered by structure, not location.
Previously, BMAD would skip the AgentVibes installer if it detected
existing hook files (.claude/hooks/bmad-speak.sh and play-tts.sh).
This prevented users from getting AgentVibes updates.
Now when user says "Yes" to AgentVibes, it always runs the installer,
ensuring users get the latest version with new features and fixes.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Generate _cfg/agent-voice-map.csv during BMAD install with voice and intro columns
- Update party mode instructions to have 3-4 agents introduce themselves on activation
- Each agent speaks their personalized intro message via TTS
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Move AgentVibes TTS prompt to after IDE/tool selection
- Default to Y if Claude Code is selected (since AgentVibes only works with Claude Code)
- Default to N if Claude Code is not selected
- Remove mid-install warning about AgentVibes not being installed (handled at end)
- Pass configured IDEs to quick update flow for consistent behavior
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
When BMAD is invoked with --prefix flag, npm sets many npm_config_*
and npm_package_* environment variables. These caused npx agentvibes
to look for files in the wrong directory (/prefix/lib instead of cwd).
Changes:
- tools/cli/commands/install.js: Filter out ALL npm_config_* and
npm_package_* env vars before calling npx agentvibes
- tools/cli/installers/lib/core/installer.js: Add AgentVibes prompt
to Quick Update flow so existing installations can enable TTS
- tools/cli/installers/lib/core/manifest-generator.js: Track
agentVibes configuration in manifest.yaml
This ensures AgentVibes runs in a clean environment regardless of
how BMAD was invoked (with or without --prefix flag).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add markdownlint-cli2 as dev dependency
- Add lint:md script to package.json
- Add markdownlint job to CI workflow
- Configure 5 rules: heading-increment, no-duplicate-heading,
no-trailing-punctuation, no-bare-urls, no-space-in-emphasis
- Fix existing violations across 19 markdown files
- No auto-fix to prevent destructive changes
Closes#1034🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Brian <bmadcode@gmail.com>
The agent build caching never worked - BUILD-META comments were
never written to output files, so every build acted like --force.
Since building all 29 agents takes ~300ms, caching provided no
meaningful benefit. Removed ~190 lines of dead code including
checkIfNeedsRebuild, checkBuildStatus, buildMetadataComment,
and the --force flag.
Co-authored-by: Brian <bmadcode@gmail.com>
* update 2024 to 2025
* fix(workflows): remove hardcoded years from WebSearch queries
Years in search queries (2024/2025) do not improve results - search
engines already prioritize current documentation. Tested all patterns
and confirmed identical quality results with/without years.
Removes years from:
- step-03-starter.md (5 queries)
- step-04-decisions.md (2 queries)
- game-architecture/instructions.md (2 queries)
Leaves file-utils.md unchanged (test fixture data, not a search query).
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(workflows): remove year placeholders from research WebSearch queries
Search engines return current results regardless of year - removes
{{current_year}} and hardcoded 2025 from step-05-technical-trends.md
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* refactor(workflows): replace {{current_year}} with semantic alternatives
Replaces year placeholder with context-appropriate wording:
- 'current data' for up-to-date information
- 'web searches' without year qualifier
- Updated failure mode to focus on using web searches
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* fix(workflows): clarify failure mode about stale training data
Rephrased to explicitly mention training data cutoff as the reason
to use web searches for current technology trends.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
* refactor(workflows): make web search references platform-agnostic
- Remove hardcoded year references from WebSearch queries
- Replace `WebSearch` tool name with natural language "search the web"
- Soften "training data is stale" to "verify and supplement your knowledge"
- Add web search prerequisite check to research workflow
- Add platform-agnostic design note to CLAUDE.md
This framework targets 15+ agentic platforms, not just Claude Code.
Tool-specific syntax like `WebSearch:` won't work across all platforms.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* refactor(research): clean up prompts and routing
---------
Co-authored-by: Pomazan Bohdan <pomazan.bogdan@gmail.com>
Co-authored-by: Brian <bmadcode@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
The manifest writer hardcoded 'bmad' as the path prefix regardless of
the actual folder name (.bmad, bmad, etc). The reader had a matching
hardcoded strip, so it worked by accident.
Now paths are stored relative to bmadDir without any prefix. Legacy
fallback strips 'bmad/' on read - safe because no real path inside
bmadDir would start with 'bmad/'.
Co-authored-by: Brian <bmadcode@gmail.com>
- Remove validate-prd workflow references from all workflow path YAML files
- Update Excalidraw diagram: remove Validate PRD box and zombie JSON elements
- Re-export SVG at 1x scale
- Standardize implementation-readiness descriptions across all docs
- Add validation script (validate-svg-changes.sh) and README for SVG export process
- Correct Excalidraw timestamps
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Brian <bmadcode@gmail.com>
- Fix checklist to only accept 'review' status (not 'ready-for-review')
- Include MEDIUM issues in done/in-progress status determination
- Initialize and track fixed_count/action_count variables for summary
- Add sprint-status.yaml sync when story status changes
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude <noreply@anthropic.com>
Replace 'looks good' with `looks good` to avoid nested single quote
issues when IDEs generate command files from workflow YAML.
Co-authored-by: Brian <bmadcode@gmail.com>
## Summary
- Track all files with TTS injection applied during installation
- Display informative summary explaining what TTS injection does
- Show backup location and restore command for recovery
## What is TTS Injection?
TTS (Text-to-Speech) injection adds voice instructions to BMAD agents,
enabling them to speak their responses aloud using AgentVibes.
Example: When you activate the PM agent, it will greet you with
spoken audio like "Hey! I'm your Project Manager. How can I help?"
## Changes
- **installer.js**: Track files in `processAgentFiles()`, `buildStandaloneAgents()`,
and `rebuildAgentFiles()` when TTS markers are processed
- **compiler.js**: Add TTS injection support for custom agent compilation
- **ui.js**: Enhanced installation summary showing:
- Explanation of what TTS injection is with example
- List of all files with TTS injection applied (grouped by type)
- Backup location (~/.bmad-tts-backups/)
- Restore command for recovery
## Example Output
```
═══════════════════════════════════════════════════
AgentVibes TTS Injection Summary
═══════════════════════════════════════════════════
What is TTS Injection?
TTS (Text-to-Speech) injection adds voice instructions to BMAD agents,
enabling them to speak their responses aloud using AgentVibes.
Example: When you activate the PM agent, it will greet you with
spoken audio like "Hey! I'm your Project Manager. How can I help?"
✅ TTS injection applied to 11 file(s):
Party Mode (multi-agent conversations):
• .bmad/core/workflows/party-mode/instructions.md
Agent TTS (individual agent voices):
• .bmad/bmm/agents/analyst.md
• .bmad/bmm/agents/architect.md
...
Backups & Recovery:
Pre-injection backups are stored in:
~/.bmad-tts-backups/
To restore original files (removes TTS instructions):
bmad-tts-injector.sh --restore /path/to/.bmad
```
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Paul Preibisch <paul@paulpreibisch.com>
Co-authored-by: Claude <noreply@anthropic.com>
Paul Preibisch
Brian Madison
Alex Verkhovsky
Wendy Smoak
Hang
Nguyen Quang Huy
Paul Preibisch