gitmoot

Coordinate AI coding agents across repositories and GitHub pull request workflows.

37|31|Updated May 20, 2026
One-click install
npx skills add https://github.com/gitmoot/gitmoot --skill gitmoot
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gitmoot
Source: https://github.com/gitmoot/gitmoot/tree/main
Command: npx skills add https://github.com/gitmoot/gitmoot --skill gitmoot

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Gitmoot coordinates AI coding agents across repositories, sessions, runtimes, and days while preserving an auditable GitHub pull request workflow and local source of truth.

Core Features & Use Cases

  • Multi-Agent Coordination: Route asking, reviewing, and implementation work across Codex, Claude Code, Kimi Code, and deterministic shell runtimes.
  • Reliable Workflow Management: Manage daemon jobs, repository checkouts, branch and runtime-session locks, budgets, crash recovery, orchestration trees, pipelines, and structured results.
  • GitHub-Centered Collaboration: Use pull request comments, issues, branches, commits, reviews, and merges as the shared surface for human and agent work.
  • Agent Customization and Optimization: Create reusable templates, capture current-chat workflows, orchestrate bounded agent teams, curate memory, and run SkillOpt evaluation and promotion workflows.
  • Use Case: A team can ask a coordinator agent to review a pull request using independent Codex, Claude, and Kimi reviewers, synthesize their findings, and return an auditable result through the existing GitHub workflow.

Quick Start

Use Gitmoot to verify the local installation, confirm GitHub CLI access, inspect the repository status, and coordinate an agent task through the appropriate runtime.

Frequently Asked Questions about gitmoot

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate multiple AI coding agents across GitHub pull request workflows?

Multi-agent orchestration routes asking, reviewing, and implementation work across Codex, Claude Code, Kimi Code, and shell runtimes. It uses pull request comments, branches, and merges as the shared surface for human and agent collaboration, returning auditable results.

Can I use different AI agent runtimes like Codex and Claude Code together in one pull request review?

Yes, agent orchestration supports routing tasks across Codex, Claude Code, Kimi Code, and deterministic shell runtimes. A coordinator agent can dispatch independent reviewers using these different runtimes, synthesize their findings, and return a unified result.

What do I need to set up before automating pull request reviews with AI agents?

You need the Gitmoot CLI, Git, authenticated GitHub CLI access, network access to GitHub, local SQLite state, and a supported agent runtime. These prerequisites ensure reliable workflow management and preserve an auditable GitHub pull request workflow.

How does multi-agent orchestration handle crash recovery and workflow management for GitHub tasks?

Workflow management handles daemon jobs, repository checkouts, branch locks, runtime-session locks, budgets, and crash recovery. It maintains local SQLite state and orchestration trees to preserve an auditable source of truth across sessions, repositories, and days.

What is the best way to manage AI agent memory and templates for reusable GitHub workflows?

Agent customization creates reusable templates, captures current-chat workflows, curates memory, and runs SkillOpt evaluation workflows. This allows teams to optimize bounded agent teams and promote effective configurations for recurring pull request automation tasks.

Are there limitations when coordinating AI agents across multiple repositories and sessions?

Coordination requires network access to GitHub and local SQLite state, meaning offline functionality is limited. Runtime-session locks and branch locks are enforced to prevent conflicts, and supported agent runtimes are restricted to Codex, Claude Code, Kimi Code, and shell environments.