GitHub Autonomous Project Management

# GitHub Autonomous Project Management Manage large projects on GitHub using both the gh CLI and browser automation. This skill focuses on coordinating big multi-step projects across multiple repos and agents. ## Core Operations ### 1. Creating Repos ```bash gh repo create <name> --public/--private --description "..." ``` ### 2. Managing Issues (Task Tracking) ```bash gh issue create --title "Task title" --body "Description" --label "enhancement" gh issue list gh issue view <number> ``` Use issues as your task tracking system — each issue is one unit of work. ### 3. Pull Requests ```bash gh pr create --title "Feature" --body "Description" --base main gh pr list gh pr view <number> gh pr merge <number> ``` ### 4. GitHub Projects (Project Boards) ```bash gh project create --title "Project Name" # Add items to project boards for tracking work across repos ``` Project boards give you a Kanban-style view across multiple repos. ### 5. Branch Management - Create feature branches for each task - One branch per agent when parallelizing work - Merge via PRs for clean integration ### 6. Agent Context Files Set up these files in every repo for agent context: - **CLAUDE.md** — instructions for Claude Code sessions - **AGENTS.md** — instructions for any AI agent working on the repo ### 7. Multi-Agent Coordination Pattern for coordinating multiple agents on one repo: 1. Each agent gets its own branch 2. Agents work independently on their branches 3. PRs are created for integration 4. Review and merge to main ### 8. Advanced gh API Usage ```bash gh api repos/{owner}/{repo}/... ``` Use for operations not covered by standard gh commands. ### 9. Labels and Milestones ```bash gh label create "priority:high" --color FF0000 gh issue create --milestone "v1.0" ``` Organize work with labels and milestones. ## Autonomous Work Pattern The key pattern for autonomous project management: 1. **Create an issue** describing the task 2. **Create a branch** for the work 3. **Dispatch to an agent** (Claude Code, Codex, etc.) 4. Agent works and **creates a PR** 5. **Review and merge** This pattern scales to any number of parallel tasks. ## CLI vs Browser The gh CLI is installed and authenticated on the workstation. **Prefer CLI over browser automation for GitHub** — it is more reliable and scriptable. Use browser automation only for features not available via CLI. ## No Project System Warning These platforms have NO project system or persistent memory. Agents must include full context in every prompt: - Who the user is and their account details - What they are working on and why - The complete goal and any relevant history - All necessary URLs, credentials references, and prior results Never assume the agent remembers anything from previous interactions. ## Running on Mac Minis These browser automation tasks run on Mac Minis via mac-control on port 8200: - **Vault**: 192.168.86.27:8200 - **Bots**: 192.168.86.50:8200 - **Jarvis**: 192.168.86.51:8200 Use the mac-control API to send browser automation commands. Example: ``` POST http://192.168.86.27:8200/execute ```