aigroup-workflow 1.0.0 → 1.1.0

package/skills/jarvis/engineering-team/templates/agent-template.md

@@ -1,318 +0,0 @@
----
-name: cs-agent-name
-description: One-line description of what this agent does (keep under 150 characters)
-skills: skill-folder-name
-domain: domain-name
-model: sonnet
-tools: [Read, Write, Bash, Grep, Glob]
----
-
-# Agent Name
-
-<!--
-  INSTRUCTIONS FOR USING THIS TEMPLATE:
-
-  1. Replace "cs-agent-name" with your agent's name (use kebab-case with cs- prefix)
-  2. Replace "Agent Name" with the display name (Title Case)
-  3. Fill in all sections below following the structure
-  4. Test all relative paths (../../) before committing
-  5. Ensure minimum 3 workflows documented
-  6. Provide concrete integration examples
-  7. Define measurable success metrics
-
-  EXAMPLES OF COMPLETED AGENTS:
-  - agents/marketing/cs-content-creator.md
-  - agents/marketing/cs-demand-gen-specialist.md
-  - agents/c-level/cs-ceo-advisor.md
-  - agents/c-level/cs-cto-advisor.md
-  - agents/product/cs-product-manager.md
--->
-
-## Purpose
-
-<!--
-  Write 2-3 paragraphs describing:
-  - What this agent does
-  - Who it's designed for (target users)
-  - How it enables better decisions/outcomes
-  - The specific gap it bridges
-
-  Example structure:
-  Paragraph 1: Agent's primary function and skill orchestration
-  Paragraph 2: Target audience and their pain points
-  Paragraph 3: Value proposition and outcome focus
--->
-
-[Paragraph 1: Describe what this agent does and which skill package it orchestrates]
-
-[Paragraph 2: Describe target users, their roles, and why they need this agent]
-
-[Paragraph 3: Explain the gap this agent bridges and the outcomes it enables]
-
-## Skill Integration
-
-**Skill Location:** `../../domain-skill/skill-name/`
-
-<!--
-  Document how this agent integrates with the underlying skill package.
-  Test all paths to ensure they resolve correctly from agents/domain/ directory.
--->
-
-### Python Tools
-
-<!--
-  List all Python automation tools from the skill package.
-  Minimum 1 tool, ideally 2-4 tools.
-
-  For each tool, provide:
-  - Clear purpose statement
-  - Exact file path (relative from agent location)
-  - Usage examples with arguments
-  - Key features
-  - Common use cases
--->
-
-1. **Tool Name**
-   - **Purpose:** What this tool does (one sentence)
-   - **Path:** `../../domain-skill/skill-name/scripts/tool_name.py`
-   - **Usage:** `python ../../domain-skill/skill-name/scripts/tool_name.py [arguments]`
-   - **Features:** Key capabilities (bullet list)
-   - **Use Cases:** When to use this tool
-
-2. **Second Tool** (if applicable)
-   - **Purpose:** What this tool does
-   - **Path:** `../../domain-skill/skill-name/scripts/second_tool.py`
-   - **Usage:** `python ../../domain-skill/skill-name/scripts/second_tool.py [arguments]`
-   - **Features:** Key capabilities
-   - **Use Cases:** When to use this tool
-
-### Knowledge Bases
-
-<!--
-  List reference documentation from the skill package.
-  These are markdown files with frameworks, best practices, templates.
--->
-
-1. **Reference Name**
-   - **Location:** `../../domain-skill/skill-name/references/reference_file.md`
-   - **Content:** What knowledge this file contains
-   - **Use Case:** When to consult this reference
-
-2. **Second Reference** (if applicable)
-   - **Location:** `../../domain-skill/skill-name/references/second_reference.md`
-   - **Content:** What knowledge this file contains
-   - **Use Case:** When to consult this reference
-
-### Templates
-
-<!--
-  List user-facing templates from the skill package's assets/ folder.
-  Optional section - only if skill has templates.
--->
-
-1. **Template Name** (if applicable)
-   - **Location:** `../../domain-skill/skill-name/assets/template.md`
-   - **Use Case:** When users would copy and customize this template
-
-## Workflows
-
-<!--
-  Document MINIMUM 3 workflows. Ideally 4 workflows.
-  Each workflow must have: Goal, Steps, Expected Output, Time Estimate
-
-  Workflow types to consider:
-  - Primary use case (most common)
-  - Advanced use case (complex scenario)
-  - Integration use case (combining multiple tools)
-  - Automated workflow (scripting/batching)
--->
-
-### Workflow 1: [Primary Use Case Name]
-
-**Goal:** One-sentence description of what this workflow accomplishes
-
-**Steps:**
-1. **[Action Step]** - Description of first step
-   ```bash
-   # Command example if applicable
-   python ../../domain-skill/skill-name/scripts/tool.py input.txt
-   ```
-2. **[Action Step]** - Description of second step
-3. **[Action Step]** - Description of third step
-4. **[Action Step]** - Description of fourth step
-5. **[Action Step]** - Description of final step
-
-**Expected Output:** What success looks like (deliverable, metric, decision made)
-
-**Time Estimate:** How long this workflow typically takes
-
-**Example:**
-```bash
-# Complete workflow example with real commands
-command1
-command2
-# Review output
-```
-
-### Workflow 2: [Advanced Use Case Name]
-
-**Goal:** One-sentence description
-
-**Steps:**
-1. **[Action Step]** - Description
-2. **[Action Step]** - Description
-3. **[Action Step]** - Description
-4. **[Action Step]** - Description
-
-**Expected Output:** What success looks like
-
-**Time Estimate:** Duration estimate
-
-### Workflow 3: [Integration Use Case Name]
-
-**Goal:** One-sentence description
-
-**Steps:**
-1. **[Action Step]** - Description
-2. **[Action Step]** - Description
-3. **[Action Step]** - Description
-
-**Expected Output:** What success looks like
-
-**Time Estimate:** Duration estimate
-
-### Workflow 4: [Optional Fourth Workflow]
-
-<!-- Delete this section if you only have 3 workflows -->
-
-**Goal:** One-sentence description
-
-**Steps:**
-1. **[Action Step]** - Description
-2. **[Action Step]** - Description
-
-**Expected Output:** What success looks like
-
-**Time Estimate:** Duration estimate
-
-## Integration Examples
-
-<!--
-  Provide 2-3 concrete code examples showing real-world usage.
-  These should be copy-paste ready bash scripts or commands.
-
-  Example types:
-  - Weekly/daily automation scripts
-  - Multi-tool workflows
-  - Output processing examples
-  - Real-time monitoring
--->
-
-### Example 1: [Example Name]
-
-```bash
-#!/bin/bash
-# script-name.sh - Brief description
-
-# Setup variables
-INPUT_FILE=$1
-
-# Execute workflow
-python ../../domain-skill/skill-name/scripts/tool.py "$INPUT_FILE"
-
-# Process output
-echo "Analysis complete. Review results above."
-```
-
-### Example 2: [Example Name]
-
-```bash
-# Multi-step workflow example
-
-# Step 1: Prepare data
-echo "Step 1: Data preparation"
-
-# Step 2: Run analysis
-python ../../domain-skill/skill-name/scripts/tool.py input.csv
-
-# Step 3: Generate report
-echo "Report generation complete"
-```
-
-### Example 3: [Example Name]
-
-```bash
-# Automation example (e.g., weekly report, daily check)
-
-DATE=$(date +%Y-%m-%d)
-echo "📊 Report for $DATE"
-
-# Execute tools
-python ../../domain-skill/skill-name/scripts/tool.py current-data.csv > report-$DATE.txt
-```
-
-## Success Metrics
-
-<!--
-  Define how to measure this agent's effectiveness.
-  Group metrics into logical categories (3-4 categories).
-  Each metric should be specific and measurable.
-
-  Categories might include:
-  - Quality metrics
-  - Efficiency metrics
-  - Business impact metrics
-  - User satisfaction metrics
--->
-
-**[Metric Category 1]:**
-- **[Metric Name]:** Target value or improvement percentage
-- **[Metric Name]:** Target value or improvement percentage
-- **[Metric Name]:** Target value or improvement percentage
-
-**[Metric Category 2]:**
-- **[Metric Name]:** Target value or improvement percentage
-- **[Metric Name]:** Target value or improvement percentage
-
-**[Metric Category 3]:**
-- **[Metric Name]:** Target value or improvement percentage
-- **[Metric Name]:** Target value or improvement percentage
-
-**[Metric Category 4]** (optional):
-- **[Metric Name]:** Target value or improvement percentage
-
-## Related Agents
-
-<!--
-  Cross-reference other agents in the same domain or related domains.
-  Use relative paths from agents/ directory.
-  Explain how agents complement each other.
--->
-
-- [cs-related-agent](../domain/cs-related-agent.md) - How this agent relates (e.g., "Provides strategic context for tactical execution")
-- [cs-another-agent](cs-another-agent.md) - How this agent relates (same directory)
-- [cs-future-agent](cs-future-agent.md) - Planned agent (mark as "planned")
-
-## References
-
-<!--
-  Link to all related documentation.
-  Always include these three links with correct relative paths.
--->
-
-- **Skill Documentation:** [../../domain-skill/skill-name/SKILL.md](../../domain-skill/skill-name/SKILL.md)
-- **Domain Guide:** [../../domain-skill/CLAUDE.md](../../domain-skill/CLAUDE.md)
-- **Agent Development Guide:** [../CLAUDE.md](../CLAUDE.md)
-
----
-
-<!--
-  Update metadata when publishing.
-  Sprint format: sprint-MM-DD-YYYY
-  Status: Production Ready, Beta, Alpha
--->
-
-**Last Updated:** [Date]
-**Sprint:** [sprint-MM-DD-YYYY] (Day X)
-**Status:** Production Ready
-**Version:** 1.0

package/skills/max/ccpm/AGENTS.md

@@ -1,114 +0,0 @@
-# Agents
-
-**[中文文档 (Chinese Documentation)](doc/AGENTS_ZH.md)**
-
-Specialized agents that do heavy work and return concise summaries to preserve context.
-
-## Core Philosophy
-
-> “Don't anthropomorphize subagents. Use them to organize your prompts and elide context. Subagents are best when they can do lots of work but then provide small amounts of information back to the main conversation thread.”
->
-> – Adam Wolff, Anthropic
-
-## Available Agents
-
-### 🔍 `code-analyzer`
-- **Purpose**: Hunt bugs across multiple files without polluting main context
-- **Pattern**: Search many files → Analyze code → Return bug report
-- **Usage**: When you need to trace logic flows, find bugs, or validate changes
-- **Returns**: Concise bug report with critical findings only
-
-### 📄 `file-analyzer`
-- **Purpose**: Read and summarize verbose files (logs, outputs, configs)
-- **Pattern**: Read files → Extract insights → Return summary
-- **Usage**: When you need to understand log files or analyze verbose output
-- **Returns**: Key findings and actionable insights (80-90% size reduction)
-
-### 🧪 `test-runner`
-- **Purpose**: Execute tests without dumping output to main thread
-- **Pattern**: Run tests → Capture to log → Analyze results → Return summary
-- **Usage**: When you need to run tests and understand failures
-- **Returns**: Test results summary with failure analysis
-
-### 🔀 `parallel-worker`
-- **Purpose**: Coordinate multiple parallel work streams for an issue
-- **Pattern**: Read analysis → Spawn sub-agents → Consolidate results → Return summary
-- **Usage**: When executing parallel work streams in a worktree
-- **Returns**: Consolidated status of all parallel work
-
-## Why Agents?
-
-Agents are **context firewalls** that protect the main conversation from information overload:
-
-```
-Without Agent:
-Main thread reads 10 files → Context explodes → Loses coherence
-
-With Agent:
-Agent reads 10 files → Main thread gets 1 summary → Context preserved
-```
-
-## How Agents Preserve Context
-
-1. **Heavy Lifting** - Agents do the messy work (reading files, running tests, implementing features)
-2. **Context Isolation** - Implementation details stay in the agent, not the main thread
-3. **Concise Returns** - Only essential information returns to main conversation
-4. **Parallel Execution** - Multiple agents can work simultaneously without context collision
-
-## Example Usage
-
-```bash
-# Analyzing code for bugs
-Task: "Search for memory leaks in the codebase"
-Agent: code-analyzer
-Returns: "Found 3 potential leaks: [concise list]"
-Main thread never sees: The hundreds of files examined
-
-# Running tests
-Task: "Run authentication tests"
-Agent: test-runner
-Returns: "2/10 tests failed: [failure summary]"
-Main thread never sees: Verbose test output and logs
-
-# Parallel implementation
-Task: "Implement issue #1234 with parallel streams"
-Agent: parallel-worker
-Returns: "Completed 4/4 streams, 15 files modified"
-Main thread never sees: Individual implementation details
-```
-
-## Creating New Agents
-
-New agents should follow these principles:
-
-1. **Single Purpose** - Each agent has one clear job
-2. **Context Reduction** - Return 10-20% of what you process
-3. **No Roleplay** - Agents aren't "experts", they're task executors
-4. **Clear Pattern** - Define input → processing → output pattern
-5. **Error Handling** - Gracefully handle failures and report clearly
-
-## Anti-Patterns to Avoid
-
-❌ **Creating "specialist" agents** (database-expert, api-expert)
-   Agents don't have different knowledge - they're all the same model
-
-❌ **Returning verbose output**
-   Defeats the purpose of context preservation
-
-❌ **Making agents communicate with each other**
-   Use a coordinator agent instead (like parallel-worker)
-
-❌ **Using agents for simple tasks**
-   Only use agents when context reduction is valuable
-
-## Integration with PM System
-
-Agents integrate seamlessly with the PM command system:
-
-- `/pm:issue-analyze` → Identifies work streams
-- `/pm:issue-start` → Spawns parallel-worker agent
-- parallel-worker → Spawns multiple sub-agents
-- Sub-agents → Work in parallel in the worktree
-- Results → Consolidated back to main thread
-
-This creates a hierarchy that maximizes parallelism while preserving context at every level.

package/skills/max/ccpm/CHANGELOG.md

@@ -1,111 +0,0 @@
-# CCPM Changelog
-
-## [2025-01-24] - Major Cleanup & Issue Resolution Release
-
-### 🎯 Overview
-Resolved 10 of 12 open GitHub issues, modernized command syntax, improved documentation, and enhanced system accuracy. This release focuses on stability, usability, and addressing community feedback.
-
-### ✨ Added
-- **Local Mode Support** ([#201](https://github.com/automazeio/ccpm/issues/201))
-  - Created `LOCAL_MODE.md` with comprehensive offline workflow guide
-  - All core commands (prd-new, prd-parse, epic-decompose) work without GitHub
-  - Clear distinction between local-only vs GitHub-dependent commands
-
-- **Automatic GitHub Label Creation** ([#544](https://github.com/automazeio/ccpm/issues/544))
-  - Enhanced `init.sh` to automatically create `epic` and `task` labels
-  - Proper colors: `epic` (green #0E8A16), `task` (blue #1D76DB)  
-  - Eliminates manual label setup during project initialization
-
-- **Context Creation Accuracy Safeguards** ([#48](https://github.com/automazeio/ccpm/issues/48))
-  - Added mandatory self-verification checkpoints in context commands
-  - Implemented evidence-based analysis requirements
-  - Added uncertainty flagging with `⚠️ Assumption - requires verification`
-  - Enhanced both `/context:create` and `/context:update` with accuracy validation
-
-### 🔄 Changed
-- **Modernized Command Syntax** ([#531](https://github.com/automazeio/ccpm/issues/531))
-  - Updated 14 PM command files to use concise `!bash` execution pattern
-  - Simplified `allowed-tools` frontmatter declarations
-  - Reduced token usage and improved Claude Code compatibility
-
-- **Comprehensive README Overhaul** ([#323](https://github.com/automazeio/ccpm/issues/323))
-  - Clarified PRD vs Epic terminology and definitions
-  - Streamlined workflow explanations and removed redundant sections
-  - Fixed installation instructions and troubleshooting guidance
-  - Improved overall structure and navigation
-
-### 📋 Research & Community Engagement
-- **Multi-Tracker Support Analysis** ([#200](https://github.com/automazeio/ccpm/issues/200))
-  - Researched CLI availability for Linear, Trello, Azure DevOps, Jira
-  - Identified Linear as best first alternative to GitHub Issues
-  - Provided detailed implementation roadmap for future development
-
-- **GitLab Support Research** ([#588](https://github.com/automazeio/ccpm/issues/588))  
-  - Confirmed strong `glab` CLI support for GitLab integration
-  - Invited community contributor to submit existing GitLab implementation as PR
-  - Updated project roadmap to include GitLab as priority platform
-
-### 🐛 Clarified Platform Limitations
-- **Windows Shell Compatibility** ([#609](https://github.com/automazeio/ccpm/issues/609))
-  - Documented as Claude Code platform limitation (requires POSIX shell)
-  - Provided workarounds and alternative solutions
-
-- **Codex CLI Integration** ([#585](https://github.com/automazeio/ccpm/issues/585))
-  - Explained future multi-AI provider support in new CLI architecture
-
-- **Parallel Worker Agent Behavior** ([#530](https://github.com/automazeio/ccpm/issues/530))
-  - Clarified agent role as coordinator, not direct coder
-  - Provided implementation guidance and workarounds
-
-### 🔒 Security
-- **Privacy Documentation Fix** ([#630](https://github.com/automazeio/ccpm/issues/630))
-  - Verified resolution via PR #631 (remove real repository references)
-
-### 💡 Proposed Features
-- **Bug Handling Workflow** ([#654](https://github.com/automazeio/ccpm/issues/654))
-  - Designed `/pm:attach-bug` command for automated bug tracking
-  - Proposed lightweight sub-issue integration with existing infrastructure
-  - Community feedback requested on implementation approach
-
-### 📊 Issues Resolved
-**Closed**: 10 issues  
-**Active Proposals**: 1 issue (#654)  
-**Remaining Open**: 1 issue (#653)
-
-#### Closed Issues:
-- [#630](https://github.com/automazeio/ccpm/issues/630) - Privacy: Remove real repo references ✅  
-- [#609](https://github.com/automazeio/ccpm/issues/609) - Windows shell error (platform limitation) ✅
-- [#585](https://github.com/automazeio/ccpm/issues/585) - Codex CLI compatibility (architecture update) ✅  
-- [#571](https://github.com/automazeio/ccpm/issues/571) - Figma MCP support (platform feature) ✅
-- [#531](https://github.com/automazeio/ccpm/issues/531) - Use !bash in custom slash commands ✅
-- [#323](https://github.com/automazeio/ccpm/issues/323) - Improve README.md ✅
-- [#201](https://github.com/automazeio/ccpm/issues/201) - Local-only mode support ✅
-- [#200](https://github.com/automazeio/ccpm/issues/200) - Multi-tracker support research ✅  
-- [#588](https://github.com/automazeio/ccpm/issues/588) - GitLab support research ✅
-- [#48](https://github.com/automazeio/ccpm/issues/48) - Context creation inaccuracies ✅
-- [#530](https://github.com/automazeio/ccpm/issues/530) - Parallel worker coding operations ✅
-- [#544](https://github.com/automazeio/ccpm/issues/544) - Auto-create labels during init ✅
-- [#947](https://github.com/automazeio/ccpm/issues/947) - Project roadmap update ✅
-
-### 🛠️ Technical Details
-- **Files Modified**: 16 core files + documentation
-- **New Files**: `LOCAL_MODE.md`, `CONTEXT_ACCURACY.md`  
-- **Commands Updated**: All 14 PM slash commands modernized
-- **Backward Compatibility**: Fully maintained
-- **Dependencies**: No new external dependencies added
-
-### 🏗️ Project Health
-- **Issue Resolution Rate**: 83% (10/12 issues closed)
-- **Documentation Coverage**: Significantly improved
-- **Community Engagement**: Active contributor invitation and feedback solicitation
-- **Code Quality**: Enhanced accuracy safeguards and validation
-
-### 🚀 Next Steps
-1. Community feedback on bug handling proposal (#654)
-2. GitLab integration PR review and merge
-3. Linear platform integration (pending demand)
-4. Enhanced testing and validation workflows
-
----
-
-*This release represents a major stability and usability milestone for CCPM, addressing the majority of outstanding community issues while establishing a foundation for future multi-platform support.*

package/skills/max/ccpm/COMMANDS.md

@@ -1,159 +0,0 @@
-# Commands
-
-**[中文文档 (Chinese Documentation)](doc/COMMANDS_ZH.md)**
-
-Complete reference of all commands available in the Claude Code PM system.
-
-> **Note**: Project Management commands (`/pm:*`) are documented in the main [README.md](README.md#command-reference).
-
-## Table of Contents
-
-- [Context Commands](#context-commands)
-- [Testing Commands](#testing-commands)
-- [Utility Commands](#utility-commands)
-- [Review Commands](#review-commands)
-
-## Context Commands
-
-Commands for managing project context in `.claude/context/`.
-
-### `/context:create`
-- **Purpose**: Create initial project context documentation
-- **Usage**: `/context:create`
-- **Description**: Analyzes the project structure and creates comprehensive baseline documentation in `.claude/context/`. Includes project overview, architecture, dependencies, and patterns.
-- **When to use**: At project start or when context needs full rebuild
-- **Output**: Multiple context files covering different aspects of the project
-
-### `/context:update`
-- **Purpose**: Update existing context with recent changes
-- **Usage**: `/context:update`
-- **Description**: Refreshes context documentation based on recent code changes, new features, or architectural updates. Preserves existing context while adding new information.
-- **When to use**: After significant changes or before major work sessions
-- **Output**: Updated context files with change tracking
-
-### `/context:prime`
-- **Purpose**: Load context into current conversation
-- **Usage**: `/context:prime`
-- **Description**: Reads all context files and loads them into the current conversation's memory. Essential for maintaining project awareness.
-- **When to use**: At the start of any work session
-- **Output**: Confirmation of loaded context
-
-## Testing Commands
-
-Commands for test configuration and execution.
-
-### `/testing:prime`
-- **Purpose**: Configure testing setup
-- **Usage**: `/testing:prime`
-- **Description**: Detects and configures the project's testing framework, creates testing configuration, and prepares the test-runner agent.
-- **When to use**: Initial project setup or when testing framework changes
--  **Output**: `.claude/testing-config.md` with test commands and patterns
-
-### `/testing:run`
-- **Purpose**: Execute tests with intelligent analysis
-- **Usage**: `/testing:run [test_target]`
-- **Description**: Runs tests using the test-runner agent which captures output to logs and returns only essential results to preserve context.
-- **Options**:
-   - No arguments: Run all tests
-   - File path: Run specific test file
-   - Pattern: Run tests matching pattern
-- **Output**: Test summary with failures analyzed, no verbose output in main thread
-
-## Utility Commands
-
-General utility and maintenance commands.
-
-### `/prompt`
-- **Purpose**: Handle complex prompts with multiple references
-- **Usage**: Write your prompt in the file, then type `/prompt`
-- **Description**: Ephemeral command for when complex prompts with numerous @ references fail in direct input. The prompt is written to the command file first, then executed.
-- **When to use**: When Claude's UI rejects complex prompts
-- **Output**: Executes the written prompt
-
-### `/re-init`
-- **Purpose**: Update or create CLAUDE.md with PM rules
-- **Usage**: `/re-init`
-- **Description**: Updates the project's CLAUDE.md file with rules from `.claude/CLAUDE.md`, ensuring Claude instances have proper instructions.
-- **When to use**: After cloning PM system or updating rules
-- **Output**: Updated CLAUDE.md in project root
-
-## Review Commands
-
-Commands for handling external code review tools.
-
-### `/code-rabbit`
-- **Purpose**: Process CodeRabbit review comments intelligently
-- **Usage**: `/code-rabbit` then paste comments
-- **Description**: Evaluates CodeRabbit suggestions with context awareness, accepting valid improvements while ignoring context-unaware suggestions. Spawns parallel agents for multi-file reviews.
-- **Features**:
-   - Understands CodeRabbit lacks full context
-   - Accepts: Real bugs, security issues, resource leaks
-   - Ignores: Style preferences, irrelevant patterns
-   - Parallel processing for multiple files
-- **Output**: Summary of accepted/ignored suggestions with reasoning
-
-## Command Patterns
-
-All commands follow consistent patterns:
-
-### Allowed Tools
-Each command specifies its required tools in frontmatter:
-- `Read, Write, LS` - File operations
-- `Bash` - System commands
-- `Task` - Sub-agent spawning
-- `Grep` - Code searching
-
-### Error Handling
-Commands follow fail-fast principles:
-- Check prerequisites first
-- Clear error messages with solutions
-- Never leave partial state
-
-### Context Preservation
-Commands that process lots of information:
-- Use agents to shield main thread from verbose output
-- Return summaries, not raw data
-- Preserve only essential information
-
-## Creating Custom Commands
-
-To add new commands:
-
-1. **Create file**: `commands/category/command-name.md`
-2. **Add frontmatter**:
-   ```yaml
-   ---
-   allowed-tools: Read, Write, LS
-   ---
-   ```
-3. **Structure content**:
-   - Purpose and usage
-   - Preflight checks
-   - Step-by-step instructions
-   - Error handling
-   - Output format
-
-4. **Follow patterns**:
-   - Keep it simple (no over-validation)
-   - Fail fast with clear messages
-   - Use agents for heavy processing
-   - Return concise output
-
-## Integration with Agents
-
-Commands often use agents for heavy lifting:
-
-- **test-runner**: Executes tests, analyzes results
-- **file-analyzer**: Summarizes verbose files
-- **code-analyzer**: Hunts bugs across codebase
-- **parallel-worker**: Coordinates parallel execution
-
-This keeps the main conversation context clean while doing complex work.
-
-## Notes
-
-- Commands are markdown files interpreted as instructions
-- The `/` prefix triggers command execution
-- Commands can spawn agents for context preservation
-- All PM commands (`/pm:*`) are documented in the main README
-- Commands follow rules defined in `/rules/`