Use when user wants to inventory autonomy branches with custom sorting, grouping, or filtering
Lists all autonomy branches with custom sorting, grouping, and filtering. Use when user runs `/list-branches` or wants to inventory branches by recency, status, or date range.
/plugin marketplace add tilmon-engineering/claude-skills/plugin install autonomy@tilmon-eng-skillsThis skill inherits all available tools. When active, it can use any tool Claude has access to.
Display inventory of all autonomy branches with user-specified sorting, grouping, and information display using computational analysis.
Core principle: Use branch-analyzer agent with Python scripts for precise analysis. Never "eyeball it".
Use this skill when:
/list-branches commandDO NOT use for:
| Step | Action | Tool |
|---|---|---|
| 1. Parse query | Extract sorting/grouping/filter requirements | Manual |
| 2. Dispatch agent | Send query to branch-analyzer | Task |
| 3. Format results | Present markdown table to user | Direct output |
Extract requirements from user's query (if provided):
Default (no query):
Parse user query for:
Example queries:
"sort by most recent, show only active"
→ Sort: recency, Filter: status=active, Show: default fields
"group by status, show metrics"
→ Group: status, Sort: recency within groups, Show: + metrics
"show branches updated in last 30 days"
→ Filter: date > (today - 30 days), Sort: recency, Show: default
Dispatch the branch-analyzer agent with detailed instructions:
Task tool with subagent_type: "autonomy:branch-analyzer"
Model: haiku
Prompt: "List all autonomy branches and analyze their status.
User query: [user's query or 'default: sort by most recent']
Requirements:
- Find all branches matching 'autonomy/*'
- For each branch, find most recent journal commit (starts with 'journal: ')
- Parse commit message for: status, metrics, blockers, next steps
- [Apply sorting: {criterion}]
- [Apply grouping: {field}]
- [Apply filtering: {criteria}]
- Generate Python script to process data
- Output markdown table with columns: [requested fields]
Use computational methods (Python scripts), do not eyeball the analysis."
Agent will:
git branch -a | grep 'autonomy/' to list all autonomy branchesDisplay agent's output to user.
Example output format:
# Autonomy Branches
Showing 3 branches (sorted by most recent update)
| Branch | Latest Iteration | Last Updated | Status | Metrics | Next |
|--------|------------------|--------------|--------|---------|------|
| experiment-a | 0028 | 2026-01-02 | blocked | MRR: $62k (+12%) | Resolve Stripe API integration |
| experiment-b | 0015 | 2025-12-28 | active | Build: 3.2min (-40%) | Implement checkout flow |
| initial-strategy | 0042 | 2025-12-15 | concluded | Churn: 8% (from 13%) | Goal achieved |
If no autonomy branches found:
No autonomy branches found.
To create your first autonomy branch:
1. Run `/create-goal` to set up an open-ended goal
2. Run `/fork-iteration <strategy-name>` to create autonomy branch
3. Run `/start-iteration` to begin work
This skill ONLY operates on autonomy/* branches:
autonomy/ prefixmain, develop)/review-progressDO NOT:
DO:
User queries are free-text and flexible:
All analysis happens via git commands:
git log <branch>| Mistake | Reality |
|---|---|
| "I'll manually list branches from git branch output" | NO. Dispatch branch-analyzer agent for computational analysis. |
| "Only 3 branches, I can eyeball the sorting" | NO. Always use Python scripts for precision. |
| "User query is unclear, I'll guess" | NO. Use AskUserQuestion to clarify if ambiguous. |
| "I'll check out each branch to read journals" | NO. Use git log to read commit messages without checkout. |
| "Non-autonomy branch appeared, I'll include it" | NO. Only autonomy/* branches. Strict filtering. |
Once branches are listed:
/branch-status <branch-name>/fork-iteration <iteration> <strategy-name>This skill should be used when the user asks to "create a hookify rule", "write a hook rule", "configure hookify", "add a hookify rule", or needs guidance on hookify rule syntax and patterns.
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.