agileflow 2.36.2 → 2.37.0

package/src/core/commands/babysit.md

@@ -2,7 +2,7 @@
 description: Interactive mentor for end-to-end feature implementation
 ---
 
-# /AgileFlow:babysit
+# /agileflow:babysit
 
 End-to-end mentor for implementing features (search stories, consult/create research, create missing pieces, guide implementation; can run commands).
 
@@ -10,6 +10,27 @@ End-to-end mentor for implementing features (search stories, consult/create rese
 
 ROLE: Babysitter (Mentor + Orchestrator)
 
+🔴 MANDATORY: AskUserQuestion FOR DECISIONS 🔴
+
+**USE AskUserQuestion for user decisions, NOT for routine operations.**
+
+**ALWAYS use AskUserQuestion for:**
+1. **INITIAL TASK SELECTION** - After loading context, present task options. Don't assume what the user wants to work on.
+2. **CHOOSING BETWEEN APPROACHES** - When 2+ valid implementation paths exist, let user decide.
+3. **END OF RESPONSE** - Always end with next step options so user can guide direction.
+4. **ARCHITECTURAL DECISIONS** - Database schema choices, API design patterns, etc.
+5. **SCOPE CLARIFICATION** - When requirements are ambiguous.
+
+**DON'T use AskUserQuestion for:**
+- Routine file writes (just do them)
+- Running commands (just run them)
+- Spawning agents (just spawn them)
+- Obvious next steps with only one sensible path
+
+**Principle**: Be helpful, not annoying. Ask for decisions, not permissions.
+
+---
+
 TODO LIST TRACKING
 **CRITICAL**: Immediately create a todo list using TodoWrite tool to track mentoring workflow:
 ```
@@ -35,7 +56,7 @@ GOAL
   3) Analyze architecture and plan small steps; assess implementation approach; propose file changes; write code & tests safely.
   4) Update docs/09-agents/status.json and bus/log.jsonl (valid JSON).
   5) Prepare PR description and next actions.
-  6) Integrate research (docs/10-research); if gaps exist, suggest /AgileFlow:context MODE=research and save results.
+  6) Integrate research (docs/10-research); if gaps exist, suggest /agileflow:context MODE=research and save results.
   7) Ensure minimal CI exists; offer to create/update .github/workflows/ci.yml or fix it if failing.
 
 🔴 ⚠️ MANDATORY CONTEXT LOADING ON FIRST RUN ⚠️ 🔴
@@ -128,7 +149,7 @@ SUGGESTIONS ENGINE
 
 **Use AskUserQuestion to present intelligent recommendations**:
 
-After loading context, analyze status.json, roadmap, and README TODOs to generate 3-7 ranked suggestions.
+After loading context, analyze status.json, roadmap, and README TODOs to generate 3-7 ranked suggestions and present them via AskUserQuestion.
 
 **Ranking Algorithm** (what a real developer/team would prioritize):
 1. **READY stories** - All acceptance criteria complete, tests stubbed, no blockers → HIGHEST PRIORITY
@@ -161,11 +182,11 @@ After loading context, analyze status.json, roadmap, and README TODOs to generat
     },
     {
       "label": "Research: JWT best practices",
-      "description": "📚 Prep work for Auth epic | Save time debugging later | /AgileFlow:context MODE=research"
+      "description": "📚 Prep work for Auth epic | Save time debugging later | /agileflow:context MODE=research"
     },
     {
       "label": "Create new epic/story",
-      "description": "💡 Start something new | Use /AgileFlow:epic or /AgileFlow:story"
+      "description": "💡 Start something new | Use /agileflow:epic or /agileflow:story"
     },
     {
       "label": "Other",
@@ -184,11 +205,11 @@ After loading context, analyze status.json, roadmap, and README TODOs to generat
 - **Limit to 5-6 options** - More than 6 creates decision paralysis
 - **"Other" is always last** - Custom input option
 
-**If research is missing/outdated**: Include research option with tip about /AgileFlow:context MODE=research
+**If research is missing/outdated**: Include research option with tip about /agileflow:context MODE=research
 
 RESEARCH INTEGRATION
 
-**💡 TIP: `/AgileFlow:context MODE=research` CAN HELP AVOID DEBUGGING HEADACHES**
+**💡 TIP: `/agileflow:context MODE=research` CAN HELP AVOID DEBUGGING HEADACHES**
 
 Consider using MODE=research when implementing new features - it can help you avoid common pitfalls, security issues, and save time debugging by learning from best practices upfront.
 
@@ -208,7 +229,7 @@ Consider using MODE=research when implementing new features - it can help you av
 
 **INTEGRATION WITH EXISTING RESEARCH**:
 - If a relevant note exists in docs/10-research: summarize 5–8 bullets + path; apply caveats to the plan.
-- If none/stale (>90 days)/conflicting: **IMMEDIATELY** propose /AgileFlow:context MODE=research TOPIC="..."; after the user pastes results, offer to save:
+- If none/stale (>90 days)/conflicting: **IMMEDIATELY** propose /agileflow:context MODE=research TOPIC="..."; after the user pastes results, offer to save:
   - docs/10-research/<YYYYMMDD>-<slug>.md (Title, Summary, Key Findings, Steps, Risks, Sources) and update docs/10-research/README.md.
 
 DEFINITION OF READY
@@ -217,7 +238,7 @@ DEFINITION OF READY
 - ✓ Test stub at docs/07-testing/test-cases/<US_ID>.md
 - ✓ Dependencies resolved and documented
 - ✓ Previous Story Insights included (if applicable to epic)
-- ✓ Story validated via `/AgileFlow:story-validate` (all checks passing)
+- ✓ Story validated via `/agileflow:story-validate` (all checks passing)
 
 ARCHITECTURE CONTEXT GUIDANCE
 
@@ -230,9 +251,12 @@ ARCHITECTURE CONTEXT GUIDANCE
 **If Architecture Context is incomplete**: Story should be "draft" not "ready"
 
 SAFE FILE OPS
-- Always show diffs; use AskUserQuestion tool to confirm before writing
+
+- Show diffs before writing so user sees what changed
 - Keep JSON valid; repair if needed (explain fix)
-- Example confirmation format:
+- For routine changes, just apply them - no need to ask
+- For significant/risky changes, use AskUserQuestion to confirm
+- Example format for risky changes:
 ```xml
 <invoke name="AskUserQuestion">
 <parameter name="questions">[{
@@ -248,12 +272,12 @@ SAFE FILE OPS
 </invoke>
 ```
 
-COMMAND EXECUTION (allowed, guarded)
-- You MAY run shell commands but only after showing the exact commands and receiving confirmation via AskUserQuestion
-- Good: list files, print snippets, run tests/linters/builds, generate scaffolds
-- Dangerous ops require explicit justification + separate confirmation
+COMMAND EXECUTION (allowed)
+
+- Run shell commands freely for: listing files, reading snippets, running tests/linters/builds, git operations
 - Capture and summarize output/errors
-- Example confirmation format:
+- For destructive operations (rm -rf, force push, etc.), use AskUserQuestion first
+- Example format for dangerous commands:
 ```xml
 <invoke name="AskUserQuestion">
 <parameter name="questions">[{
@@ -269,76 +293,117 @@ COMMAND EXECUTION (allowed, guarded)
 </invoke>
 ```
 
-AGENT SPAWNING & CONTEXT PRESERVATION (CRITICAL)
+AGENT SPAWNING - FOR COMPLEX TASKS
 
-**YOU SHOULD SPAWN AGENTS LIBERALLY** - Using agents is BETTER than doing everything yourself because:
-1. **Preserves Context**: Each agent gets a fresh context window for its specialty
-2. **Better Focus**: Agent stays focused on single task without token overhead
-3. **Parallel Work**: Multiple agents can work simultaneously
-4. **Specialized Knowledge**: Each agent has deep prompting for its domain
-5. **Cleaner Handoff**: Results come back cleanly for next step
+**USE AGENTS FOR COMPLEX WORK, HANDLE SIMPLE TASKS YOURSELF**
 
-**HOW TO SPAWN AGENTS**:
+**SPAWN AGENTS when:**
+- Task spans multiple files or modules
+- Task requires deep domain expertise (security, performance, database design)
+- Task involves significant implementation (new features, major refactors)
+- Multiple independent workstreams can run in parallel
+- Task would benefit from focused specialist attention
+
+**HANDLE YOURSELF when:**
+- Simple edits (fix a typo, add a comment, small tweaks)
+- Quick file reads or searches
+- Single-file changes with obvious implementation
+- Status updates, simple questions, coordination tasks
+- Anything that takes less effort to do than to delegate
 
-Use the **Task tool** with `subagent_type` parameter:
+**WHY USE AGENTS (when appropriate):**
+1. **Preserves Context**: Agents handle deep work, you stay lightweight
+2. **Parallel Work**: Multiple agents = faster completion
+3. **Specialist Knowledge**: Domain experts know their area deeply
+
+**HOW TO SPAWN AGENTS**:
 
 ```
 Task(
-  description: "Brief 3-5 word description of what you want",
-  prompt: "Detailed task description for the agent to execute",
+  description: "Brief 3-5 word description",
+  prompt: "Detailed task for the agent",
   subagent_type: "AgileFlow:<agent-name>"
 )
 ```
 
-**Example Usage**:
+**ASYNC AGENTS FOR PARALLEL WORK**:
+
+Use `run_in_background: true` when tasks can run in parallel:
+
 ```
-# Spawn epic-planner to create epic
+# Spawn multiple agents simultaneously
 Task(
-  description: "Create authentication epic",
-  prompt: "Create EP-0001 for user authentication system with login, logout, password reset stories",
-  subagent_type: "AgileFlow:epic-planner"
+  description: "Create API endpoint",
+  prompt: "Implement /api/users endpoint with CRUD operations",
+  subagent_type: "AgileFlow:api",
+  run_in_background: true
 )
 
-# Spawn mentor to implement feature
 Task(
-  description: "Implement user login story",
-  prompt: "Guide implementation of US-0001: User Login API endpoint with JWT auth",
-  subagent_type: "AgileFlow:mentor"
+  description: "Create user form component",
+  prompt: "Build UserForm component with validation",
+  subagent_type: "AgileFlow:ui",
+  run_in_background: true
 )
 
-# Spawn research agent for technical questions
-Task(
-  description: "Research JWT best practices",
-  prompt: "Research and document best practices for JWT token refresh, expiration, and security",
-  subagent_type: "AgileFlow:research"
-)
+# Later, collect results with TaskOutput
 ```
 
-<!-- {{AGENT_LIST}} -->
+**WHEN TO USE ASYNC AGENTS**:
+- API + UI work that can happen simultaneously
+- Research while implementation proceeds
+- Tests + documentation in parallel
+- Multiple independent stories
+- Any tasks without dependencies on each other
 
-**WHEN TO SPAWN AGENTS** (Use liberally!):
-
-**Planning Phase**:
-- Spawn epic-planner for any new feature request
-- Spawn research if tech stack is unfamiliar
-- Spawn adr-writer for major architectural decisions
-
-**Implementation Phase** (THIS IS WHERE AGENTS SHINE):
-- Spawn ui for component work (preserves your context for orchestration)
-- Spawn api for backend work (API specialist can focus deeply)
-- Spawn mentor for complex features (guide user through implementation)
-- Spawn ci for test setup (specialist in test infrastructure)
-- Spawn devops for deployment (infrastructure specialist)
+**Example Orchestration Flow**:
+```
+User: "Add user profile feature with API and UI"
+
+Babysit thinking:
+- This has API work → spawn AgileFlow:api
+- This has UI work → spawn AgileFlow:ui
+- These are independent → run in parallel!
+
+Babysit action:
+1. Spawn api agent (async) for profile endpoint
+2. Spawn ui agent (async) for profile component
+3. Monitor progress, collect results
+4. Coordinate integration
+```
 
-**Quality & Documentation Phase**:
-- Spawn adr-writer for decisions
-- Spawn research for documentation gaps
+<!-- {{AGENT_LIST}} -->
 
-**KEY PRINCIPLE**:
-- Babysit should be a LIGHTWEIGHT ORCHESTRATOR
-- Agents do the DEEP, FOCUSED WORK
-- This preserves context and improves quality
-- **Always spawn agents for domain-specific work**
+**AGENT REFERENCE** - For complex tasks:
+
+| Complex Task Type | Agent to Spawn | Async? |
+|-------------------|----------------|--------|
+| Multi-file UI feature | `AgileFlow:ui` | Yes if API also needed |
+| New API endpoints | `AgileFlow:api` | Yes if UI also needed |
+| Schema design/migrations | `AgileFlow:database` | Usually yes |
+| Test suite creation | `AgileFlow:testing` | Yes alongside impl |
+| CI/CD setup | `AgileFlow:ci` | Yes |
+| Security audit/impl | `AgileFlow:security` | Yes |
+| Performance optimization | `AgileFlow:performance` | Yes |
+| Large documentation | `AgileFlow:documentation` | Yes |
+| Technical research | `AgileFlow:research` | Yes |
+| Epic/story breakdown | `AgileFlow:epic-planner` | No (need result) |
+| Architecture decisions | `AgileFlow:adr-writer` | Yes |
+
+**PARALLEL PATTERNS** (for complex multi-domain tasks):
+- `api` + `ui` → Full-stack feature in parallel
+- `testing` + `documentation` → Quality tasks while reviewing
+- `research` + `epic-planner` → Research informs planning
+
+**SIMPLE VS COMPLEX**:
+```
+Simple (do yourself):     Complex (spawn agent):
+─────────────────────     ─────────────────────
+Fix typo in component     Build new component system
+Add one API field         Design new API module
+Write one test            Create test infrastructure
+Update README section     Document entire feature
+```
 
 ---
 
@@ -425,15 +490,18 @@ Response:
 2. Then: AG-UI for the profile page component"
 ```
 
-**WHEN NOT TO AUTO-SPAWN**:
+**WHEN NOT TO AUTO-SPAWN** (handle these yourself):
 
-- Simple questions that don't require implementation
+- Simple edits, typos, small tweaks
+- Single-file changes with obvious implementation
+- Quick questions or status checks
 - When user explicitly says "I'll do it myself"
 - When user is asking about AgileFlow itself (not their project)
+- Tasks that take less effort to do than to delegate
 
 **WHEN TO USE MULTI-EXPERT ORCHESTRATION**:
 
-For complex cross-domain tasks, use `/AgileFlow:multi-expert` instead of single agent:
+For complex cross-domain tasks, use `/agileflow:multi-expert` instead of single agent:
 
 | Scenario | Use Multi-Expert |
 |----------|------------------|
@@ -452,7 +520,7 @@ For complex cross-domain tasks, use `/AgileFlow:multi-expert` instead of single
 
 **How to invoke**:
 ```
-SlashCommand("/AgileFlow:multi-expert <question>")
+SlashCommand("/agileflow:multi-expert <question>")
 ```
 
 Or spawn directly:
@@ -493,7 +561,7 @@ CI INTEGRATION
 - On request, run tests/build/lint and summarize.
 
 IMPLEMENTATION FLOW
-1) Validate story readiness: `/AgileFlow:story-validate <STORY_ID>`
+1) Validate story readiness: `/agileflow:story-validate <STORY_ID>`
 2) Read relevant practices docs based on task type
 3) Read story's Architecture Context section FIRST
 4) Check Previous Story Insights (if not first in epic)
@@ -507,13 +575,16 @@ IMPLEMENTATION FLOW
 12) Generate PR body
 
 FIRST MESSAGE
+
 - One-line reminder of the system
 - Present intelligent suggestions using AskUserQuestion (see SUGGESTIONS ENGINE)
 - Explain: "I can also run safe commands, invoke specialized agents, and leverage auto-activating skills for templates and generators"
 
 OUTPUT
+
 - Headings, short bullets, code/diff/command blocks
-- Always end with AskUserQuestion for next action:
+- End with AskUserQuestion for next action (guide user to next decision point)
+- Example format:
 
 ```xml
 <invoke name="AskUserQuestion">

package/src/core/commands/baseline.md

@@ -5,7 +5,25 @@ argument-hint: [message] (optional)
 
 # Establish Baseline
 
-You are running the `/AgileFlow:baseline` command to create a verified checkpoint representing a known-good state of the project.
+Create a verified checkpoint representing a known-good state of the project.
+
+## Prompt
+
+ROLE: Baseline Creator
+
+INPUTS
+MESSAGE=<text>   Optional baseline message (default: "Baseline - all tests passing")
+
+ACTIONS
+1) Run pre-flight checks (session harness, tests passing, clean git)
+2) Show baseline summary (what will be included)
+3) Create git tag with timestamp (agileflow-baseline-YYYYMMDD-HHMMSS)
+4) Update environment.json (baseline_commit)
+5) Update session-state.json (baseline history)
+6) Update story frontmatter (verified_at_baseline)
+7) Display final report
+
+---
 
 ## Command Purpose
 
@@ -43,7 +61,7 @@ Mark each step complete as you finish it. This ensures nothing is forgotten duri
 
 ## Prerequisites
 
-- Session harness initialized (`/AgileFlow:session:init`)
+- Session harness initialized (`/agileflow:session:init`)
 - All tests must be passing
 - Git working tree must be clean
 - All `in_progress` stories must have `test_status: passing`
@@ -51,8 +69,8 @@ Mark each step complete as you finish it. This ensures nothing is forgotten duri
 ## Usage
 
 ```
-/AgileFlow:baseline
-/AgileFlow:baseline "Sprint 3 complete - all features tested"
+/agileflow:baseline
+/agileflow:baseline "Sprint 3 complete - all features tested"
 ```
 
 ## Execution Flow
@@ -63,7 +81,7 @@ Mark each step complete as you finish it. This ensures nothing is forgotten duri
 ```bash
 if [ ! -f "docs/00-meta/environment.json" ]; then
   echo "❌ Session harness not initialized"
-  echo "Run /AgileFlow:session:init first"
+  echo "Run /agileflow:session:init first"
   exit 1
 fi
 ```
@@ -71,7 +89,7 @@ fi
 **Check Tests:**
 ```bash
 echo "🧪 Verifying tests..."
-/AgileFlow:verify
+/agileflow:verify
 
 # Check exit code
 if [ $? -ne 0 ]; then
@@ -107,7 +125,7 @@ if [ -n "$FAILING_STORIES" ]; then
   echo "Stories with non-passing tests:"
   echo "$FAILING_STORIES"
   echo ""
-  echo "Run /AgileFlow:verify and fix failing tests first"
+  echo "Run /agileflow:verify and fix failing tests first"
   exit 1
 fi
 ```
@@ -252,7 +270,7 @@ To push baseline to remote:
 ### After Epic Completion
 ```
 User: We just finished EP-0005. All stories tested and merged.
-User: /AgileFlow:baseline "EP-0005: User authentication complete"
+User: /agileflow:baseline "EP-0005: User authentication complete"
 
 Claude: 🧪 Verifying tests...
         ✅ All tests passing (58/58)
@@ -274,7 +292,7 @@ Claude: 🧪 Verifying tests...
 ### Before Risky Refactor
 ```
 User: About to refactor the auth system. Want a safety checkpoint.
-User: /AgileFlow:baseline "Before auth refactor"
+User: /agileflow:baseline "Before auth refactor"
 
 Claude: ✅ Baseline created: agileflow-baseline-20251206-150000
 
@@ -282,13 +300,13 @@ Claude: ✅ Baseline created: agileflow-baseline-20251206-150000
           git checkout agileflow-baseline-20251206-150000
 
         Or compare tests after refactor:
-          /AgileFlow:verify
+          /agileflow:verify
           # Compare to baseline: 42/42 tests
 ```
 
 ### Sprint Checkpoint
 ```
-User: /AgileFlow:baseline "Sprint 12 end - 15 stories completed"
+User: /agileflow:baseline "Sprint 12 end - 15 stories completed"
 
 Claude: ✅ Baseline Established
 
@@ -362,7 +380,7 @@ Stories with non-passing tests:
 
 All in-progress stories must have passing tests.
 
-Run /AgileFlow:verify to check and fix tests.
+Run /agileflow:verify to check and fix tests.
 ```
 
 ### Git Not Available
@@ -396,11 +414,11 @@ Choice [1/2/3]: _
 ## Integration Points
 
 ### Called By
-- User manually via `/AgileFlow:baseline`
-- `/AgileFlow:session:init` - Creates initial baseline (optional)
+- User manually via `/agileflow:baseline`
+- `/agileflow:session:init` - Creates initial baseline (optional)
 
 ### Calls
-- `/AgileFlow:verify` - Test verification
+- `/agileflow:verify` - Test verification
 
 ### Reads
 - `docs/00-meta/environment.json` - Current baseline info
@@ -494,7 +512,7 @@ Claude: ⚠️  This will discard all changes since baseline
 ## Example Execution
 
 ```
-User: /AgileFlow:baseline "Major milestone - v1.0 ready"
+User: /agileflow:baseline "Major milestone - v1.0 ready"
 
 Claude: 🧪 Verifying all requirements...
 

package/src/core/commands/blockers.md

@@ -137,7 +137,7 @@ For each blocker, provide:
 ### Research Blockers
 - Search docs/10-research/ for related topics
 - Check for stale research (>90 days) that might need updating
-- Suggest running /AgileFlow:context MODE=research for specific topic
+- Suggest running /agileflow:context MODE=research for specific topic
 - Link to ADRs that might have context
 
 CROSS-AGENT COORDINATION ANALYSIS (v2.7.0)
@@ -280,7 +280,7 @@ US-0051 | AG-DEVOPS | Blocked: Unclear deployment target
 
   📚 Related:
     - Research: docs/10-research/20251010-deployment-comparison.md (90 days old, may be stale)
-    - Suggest: /AgileFlow:context MODE=research TOPIC="Modern deployment platforms 2025"
+    - Suggest: /agileflow:context MODE=research TOPIC="Modern deployment platforms 2025"
 
 ---
 
@@ -317,10 +317,10 @@ AG-API at WIP limit (2/2):
 4. [Low] Redistribute AG-API backlog when capacity opens
 
 Next Commands:
-- /AgileFlow:status STORY=US-0051 STATUS=ready NOTE="Clarified deployment target: Vercel"
-- /AgileFlow:context MODE=research TOPIC="Modern deployment platforms 2025"
-- /AgileFlow:validate-system (check for other inconsistencies)
-- /AgileFlow:board (visualize current state)
+- /agileflow:status STORY=US-0051 STATUS=ready NOTE="Clarified deployment target: Vercel"
+- /agileflow:context MODE=research TOPIC="Modern deployment platforms 2025"
+- /agileflow:validate-system (check for other inconsistencies)
+- /agileflow:board (visualize current state)
 ```
 
 RULES
@@ -338,7 +338,7 @@ After displaying blockers, ask:
 "Would you like me to update any blocker statuses or create resolution stories?"
 
 If yes, suggest:
-- `/AgileFlow:status STORY=<id> STATUS=ready` for resolved blockers
+- `/agileflow:status STORY=<id> STATUS=ready` for resolved blockers
 - `/story-new` for creating unblocking stories
 - `/handoff` for reassigning capacity-blocked stories
 - `/adr-new` for architectural blockers needing decisions

package/src/core/commands/board.md

@@ -223,7 +223,7 @@ INTEGRATION
 
 - Save board snapshot to docs/08-project/boards/board-<YYYYMMDD>.md
 - Track board states over time for velocity analysis
-- Optionally update /AgileFlow:velocity with latest data
+- Optionally update /agileflow:velocity with latest data
 
 WORKFLOW
 

package/src/core/commands/changelog.md

@@ -10,14 +10,20 @@ Automatically generate changelog from PR titles, commits, and tags.
 
 ROLE: Changelog Generator
 
-OBJECTIVE
-Create or update CHANGELOG.md based on merged PRs, commits, and version tags following Keep a Changelog format.
-
-INPUTS (optional)
-- VERSION=<version number> (e.g., 1.2.0) (default: auto-detect from latest tag)
-- SINCE=<tag or commit> (default: last version tag)
-- FORMAT=keepachangelog|simple|github (default: keepachangelog)
-- AUTO_COMMIT=yes|no (default: no, ask first)
+INPUTS
+VERSION=<version>   Version number (default: auto-detect from latest tag)
+SINCE=<tag>         Start point (default: last version tag)
+FORMAT=<format>     keepachangelog|simple|github (default: keepachangelog)
+AUTO_COMMIT=yes|no  Auto-commit changes (default: no)
+
+ACTIONS
+1) Detect version from latest tag if not provided
+2) Get commits/PRs since SINCE point
+3) Categorize changes (Added/Changed/Fixed/Removed)
+4) Generate changelog section in specified format
+5) Show diff and ask for confirmation
+6) Update CHANGELOG.md if approved
+7) Optionally commit the changes
 
 CHANGELOG FORMAT
 
@@ -254,7 +260,7 @@ jobs:
           fetch-depth: 0  # Get all tags
 
       - name: Generate changelog
-        run: npx claude-code /AgileFlow:generate-changelog VERSION=${{ github.event.inputs.version }}
+        run: npx claude-code /agileflow:generate-changelog VERSION=${{ github.event.inputs.version }}
 
       - name: Commit changelog
         run: |

package/src/core/commands/compress.md

@@ -6,6 +6,24 @@ description: Compress status.json by removing verbose fields and keeping only tr
 
 Reduce status.json file size by removing verbose story fields and keeping only essential tracking metadata.
 
+## Prompt
+
+ROLE: Status Compressor
+
+INPUTS
+(no arguments - operates on docs/09-agents/status.json)
+
+ACTIONS
+1) Validate status.json exists and is valid JSON
+2) Show before stats (size, lines, story count)
+3) Create backup (status.json.backup)
+4) Remove verbose fields (description, AC, architectureContext, etc.)
+5) Keep only tracking fields (story_id, status, owner, timestamps)
+6) Write compressed file
+7) Show after stats and savings
+
+---
+
 ## Purpose
 
 **Problem**: status.json exceeds 25000 tokens, causing agents to fail with "file content exceeds maximum allowed tokens"
@@ -44,7 +62,7 @@ Only essential tracking metadata remains:
 
 ## Workflow
 
-When user runs `/AgileFlow:compress`:
+When user runs `/agileflow:compress`:
 
 1. **Validate Environment**:
    - Check if `docs/09-agents/status.json` exists
@@ -109,8 +127,8 @@ When user runs `/AgileFlow:compress`:
 - After major epic completion (many stories with verbose records)
 
 **Combine with archival**:
-1. First run archival to move old completed stories: `/AgileFlow:compress` → runs `bash scripts/archive-completed-stories.sh`
-2. If still too large, run compression: `/AgileFlow:compress` → runs `bash scripts/compress-status.sh`
+1. First run archival to move old completed stories: `/agileflow:compress` → runs `bash scripts/archive-completed-stories.sh`
+2. If still too large, run compression: `/agileflow:compress` → runs `bash scripts/compress-status.sh`
 3. Result: status.json under 25000 tokens
 
 ## Safety
@@ -136,8 +154,8 @@ bash scripts/compress-status.sh
 
 ## Related Commands
 
-- `/AgileFlow:status` - View current story statuses
-- `/AgileFlow:validate` - Validate AgileFlow system health
+- `/agileflow:status` - View current story statuses
+- `/agileflow:validate` - Validate AgileFlow system health
 
 ## Example Output