home / skills / aaaaqwq / agi-super-skills / god-mode

god-mode skill

/skills/misc/god-mode

This skill provides a bird's-eye view of multi-repo projects and coaches AI agents to align instructions with commit patterns.

This is most likely a fork of the god-mode skill from openclaw
npx playbooks add skill aaaaqwq/agi-super-skills --skill god-mode

Review the files below or copy the command above to add this skill to your agents.

Files (21)
SKILL.md
4.4 KB
---
name: god-mode
description: Developer oversight and AI agent coaching. Use when viewing project status across repos, syncing GitHub data, or analyzing agents.md against commit patterns.
metadata: {"openclaw": {"requires": {"bins": ["gh", "sqlite3", "jq"]}}}
user-invocable: true
---

# god-mode Skill

> Developer oversight and AI agent coaching for OpenClaw.

## Overview

**god-mode** gives you a bird's-eye view of all your coding projects and coaches you to write better AI agent instructions.

**Key features:**
- Multi-project status dashboard
- Incremental sync from GitHub (Azure/GitLab coming)
- Agent instruction analysis based on commit patterns
- Local SQLite cache for fast queries

## Quick Start

```bash
# First-run setup
god setup

# Add a project
god projects add github:myuser/myrepo

# Sync data
god sync

# See overview
god status

# Analyze your agents.md
god agents analyze myrepo
```

## Commands

### `god status [project]`
Show overview of all projects, or details for one:
```bash
god status              # All projects
god status myproject    # One project in detail
```

### `god sync [project] [--force]`
Fetch/update data from repositories:
```bash
god sync                # Incremental sync all
god sync myproject      # Just one project
god sync --force        # Full refresh (ignore cache)
```

### `god projects`
Manage configured projects:
```bash
god projects                        # List all
god projects add github:user/repo   # Add project
god projects remove myproject       # Remove project
```

### `god agents analyze <project>`
Analyze agents.md against commit history:
```bash
god agents analyze myproject
```

Finds gaps between your agent instructions and actual work patterns, suggests improvements.

### `god agents generate <project>` (Coming Soon)
Bootstrap agents.md for a new project by analyzing repo structure.

## Configuration

Config file: `~/.config/god-mode/config.yaml`

```yaml
projects:
  - id: github:user/repo
    name: My Project      # Display name
    priority: high        # high/medium/low
    tags: [work, api]
    local: ~/code/myrepo  # Local clone path

sync:
  initialDays: 90         # First sync lookback
  commitsCacheMinutes: 60

analysis:
  agentFiles:             # Files to search for
    - agents.md
    - AGENTS.md
    - CLAUDE.md
    - .github/copilot-instructions.md
```

## Data Storage

All data stored locally in `~/.god-mode/`:
- `cache.db` - SQLite database (commits, PRs, issues, analyses)
- `contexts/` - Saved workspace contexts (v0.2)

## Authentication

god-mode uses your existing CLI authentication:

| Provider | CLI | Setup |
|----------|-----|-------|
| GitHub | `gh` | `gh auth login` |
| Azure | `az` | `az login` |
| GitLab | `glab` | `glab auth login` |

**No tokens stored by god-mode.** We delegate to CLIs you already trust.

## Requirements

- `gh` - GitHub CLI (for GitHub repos)
- `sqlite3` - Database
- `jq` - JSON processing

## Examples

### Morning Check-In
```bash
god status
# See all projects at a glance
# Notice any stale PRs or quiet projects
```

### Before Switching Projects
```bash
god status myproject
# See recent activity, open PRs, issues
# Remember where you left off
```

### Improving Your AI Assistant
```bash
god agents analyze myproject
# Get suggestions based on your actual commit patterns
# Apply recommendations to your agents.md
```

### Weekly Review
```bash
god status
# Review activity across all projects
# Identify projects needing attention
```

## Agent Workflows

### Daily Briefing (Heartbeat)
```markdown
# HEARTBEAT.md
- Run `god status` and summarize:
  - Projects with stale PRs (>3 days)
  - Projects with no activity (>5 days)
  - Open PRs needing review
```

### Agent Analysis (Cron)
```yaml
# Weekly agent instruction review
schedule: "0 9 * * 1"  # Monday 9am
task: |
  Run `god agents analyze` on high-priority projects.
  If gaps found, notify with suggestions.
```

## Troubleshooting

### "gh: command not found"
Install GitHub CLI: https://cli.github.com/

### "Not logged in to GitHub"
Run: `gh auth login`

### "No projects configured"
Add a project: `god projects add github:user/repo`

### Stale data
Force refresh: `god sync --force`

---

*OpenClaw Community Skill*  
*License: MIT*  
*Repository: https://github.com/InfantLab/god-mode-skill*

Overview

This skill provides developer oversight and AI agent coaching across multiple code projects. It aggregates commit, PR, and issue data into a local SQLite cache and surfaces status, activity trends, and instruction gaps. Use it to sync Git data, get a project-wide dashboard, and analyze agent instructions against real commit patterns.

How this skill works

The tool performs incremental syncs from GitHub (with support for other providers planned) and stores commits, PRs, and issues in a local SQLite database for fast queries. It scans configured agent instruction files (agents.md and common variants) and compares their claimed workflows to actual commit and PR patterns to identify mismatches. Command-line actions let you add projects, run syncs, view status, and run agent analysis that generates actionable suggestions.

When to use it

  • When you want a morning or weekly overview of activity across projects
  • Before switching contexts to see recent activity and open PRs/issues
  • When onboarding or reviewing AI agent instructions for a repository
  • To detect gaps between documented agent responsibilities and real work patterns
  • When preparing a status report or weekly review across multiple repos

Best practices

  • Configure projects with local clone paths for faster correlation of work and commits
  • Run regular incremental syncs to keep the cache current; use force refresh only when troubleshooting
  • Keep agent instruction files named consistently (agents.md, AGENTS.md, etc.) and included in config search list
  • Prioritize high-importance projects in config to focus automated analysis and notifications
  • Treat analysis suggestions as coach prompts—review and adapt agent docs instead of blindly applying changes

Example use cases

  • Daily check-in: run status to find stale PRs, quiet projects, and items needing review
  • Project switch: inspect a project’s recent commits, open PRs, and issue backlog before resuming work
  • Agent instruction audit: run agent analysis to surface tasks missing from agents.md and get phrasing suggestions
  • Weekly review automation: schedule weekly analyses on high-priority projects and surface suggested improvements to the team
  • Onboarding: generate a prioritized list of agent instruction gaps for a new contributor to understand project workflows

FAQ

Which authentication is required?

The skill delegates authentication to your existing CLI tools; authenticate with the provider's CLI (for example GitHub CLI) before syncing.

Where is data stored?

All data is stored locally in a hidden application directory and a SQLite cache for fast, offline queries.

Can I analyze non-GitHub repos?

GitHub is supported by default; other providers are planned and some CLIs may already work if configured.