Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills Aโ€“Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Gt Cli

ASecurity

Gas Town is a multi-agent orchestration system for Claude Code that provides persistent work tracking via git-backed hooks. Scales to 20-30 agents by eliminating context loss during restarts.

22 stars
0 votes
0 copies
0 views
Added 9/20/2026
ai-agentsgobashgitdatabase

Works with

claude codecursorcli

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add lev-os/agents --skill gt-cli --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Gt Cli?

Add the live security badge to your README โ€” it updates automatically with every re-scan.

Security grade badge for Gt Cli
[![Security: A โ€” Skills Directory](https://www.skillsdirectory.com/api/skills/lev-os-gt-cli/badge)](https://www.skillsdirectory.com/skills/lev-os-gt-cli)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
skill.md
# Gas Town (gt CLI) Skill

## Overview

Gas Town is a multi-agent orchestration system for Claude Code that provides persistent work tracking via git-backed hooks. Scales to 20-30 agents by eliminating context loss during restarts.

## When to Use

- Multi-agent orchestration with work persistence
- GUPP (propulsion) pattern for autonomous execution
- Hook-based work assignment that survives restarts
- Convoy tracking for multi-issue work
- MEOW (Mayor-Enhanced Orchestration Workflow) patterns

## CLI Command Reference

### Workspace Setup

```bash
gt install <path>              # Initialize workspace
gt rig add <name> <repo>       # Add project container
gt rig list                    # List rigs
gt crew add <name> --rig <rig> # Create crew workspace
```

### Agent Operations

```bash
gt agents                      # List active agents
gt sling <issue> <rig>         # Assign work to agent (creates hook)
gt sling <issue> --agent cursor # Override agent runtime
gt mayor attach                # Start Mayor session
gt prime                       # Startup with hook check (outputs role context)
```

### Work Tracking

```bash
gt convoy create <name> [issues...] # Create convoy for tracking
gt convoy list                      # List convoys (dashboard)
gt convoy show [id]                 # Show convoy details
gt convoy status <id>               # Status of convoy
```

### Hooks & Persistence

```bash
gt hook                        # Show hooked work
gt hook <bead-id>              # Attach work to hook
gt hook <bead-id> -f           # Force replace existing
gt hook show <agent>           # Compact view of agent's hook
gt done                        # Signal work ready for merge
gt unsling                     # Remove work from hook
```

### Molecules & Formulas

```bash
gt formula list                # List available formulas
gt formula show <name>         # Show formula details
gt formula run <name>          # Execute formula

gt mol attach <bead> <mol>     # Pin molecule to bead
gt mol detach <bead>           # Unpin molecule
gt mol status                  # Current molecule status
```

### Configuration

```bash
gt config agent set <name> <cmd>  # Set custom agent
gt config default-agent <name>    # Set default runtime
```

## GUPP Mechanism

**GUPP = Gas Town Universal Propulsion Protocol**

Core principle: "If you find something on your hook, YOU RUN IT."

### Startup Sequence (via `gt prime`)

1. Detect role from current directory
2. Check for hooked work
3. IF work hooked:
   - Display "๐Ÿšจ AUTONOMOUS WORK MODE ๐Ÿšจ"
   - Show assignment details
   - BEGIN IMMEDIATELY - no waiting
4. Send propulsion nudge via tmux
5. Run bd prime for workflow context
6. Inject pending mail

### Propulsion Loop

```bash
1. gt hook                    # What's hooked?
2. bd mol current             # Where am I in molecule?
3. Execute step
4. bd close <step> --continue # Close and auto-advance
5. GOTO 1
```

## MEOW Pattern

**MEOW = Mayor-Enhanced Orchestration Workflow**

### 7-Step Workflow

1. **Tell Mayor** - Describe what to build
2. **Mayor analyzes** - Breaks down into tasks
3. **Convoy creation** - Groups related issues
4. **Agent spawning** - Spawns polecats via `gt sling`
5. **Work distribution** - Issues hooked to agents
6. **Progress monitoring** - Track via `gt convoy status`
7. **Completion** - Mayor summarizes results

### Mayor Commands

```bash
gt mayor attach                        # Enter Mayor mode
gt convoy create "Feature X" issue-1 issue-2 --notify
gt sling issue-1 myproject             # Distribute to agent
gt convoy list                         # Monitor progress
```

## Polecat Lifecycle

### States

- `working` - Actively working on assigned issue
- `done` - Completed, ready for Witness cleanup
- `stuck` - Needs human assistance

### Commands

```bash
gt polecat spawn <name>        # Create and start polecat
gt polecat list                # List all polecats
gt polecat status <name>       # Show session info
gt polecat cycle <name>        # Graceful shutdown
gt polecat kill <name>         # Force terminate
```

## Agent Roles

| Role | Command | Responsibility |
|------|---------|----------------|
| Mayor | `gt mayor attach` | Global coordinator |
| Deacon | `gt deacon status` | Town-level health |
| Witness | `gt witness status` | Polecat health monitoring |
| Refinery | `gt refinery status` | Merge queue processing |
| Polecat | `gt polecat spawn` | Worker agent |
| Crew | `gt crew add` | Persistent worker |

## Environment Variables

```bash
GT_ROLE                # Role override
GT_RIG                 # Current rig
GT_POLECAT             # Polecat name
GT_SESSION_ID          # Session ID
BEADS_DIR              # Beads database location
```

## Anti-Patterns to Avoid

1. Don't read formulas directly - Use `bd cook` โ†’ `bd mol pour`
2. Don't wait after announcing - GUPP requires immediate execution
3. Don't use 3-command transitions - Use `bd close --continue`
4. Don't create manual convoys - Use `gt sling` (auto-creates)
5. Don't poll for work - Hooks are the assignment source

## Source

Repository: https://github.com/steveyegge/gastown

Attribution

lev-oslev-os
View sourceMore from lev-os โ†’
SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots ยท $299/mo ยท GSC-verified traffic ยท sponsors can never buy grades.

See placements

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots ยท $299/mo ยท GSC-verified traffic ยท sponsors can never buy grades.

See placements

Related Skills

Caveman

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

1023331 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3331 votes

catchup

Recovers prior coding-agent session context by running `catchup <agent> --since-compact`, which extracts a clean summary of a previous Codex, Claude Code, Antigravity, OpenCode, or Pi Agent session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", or asks to recover/summarize a previous session before continuing. Do NOT use for the current conversation, git history, or any non-agent log.

611 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants โ€” handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents โ†’