claude

claude-agent-kit

Multi-agent orchestration plugin for Claude Code with personas, planning workflows, and continuation control

0 on GitHubv0.1.0updated 8 months ago

Not in a marketplace yet, so it loads for one session with --plugin-dir.

Plugin claude-agent-kit in nicholls-inc/claude-agent-kit. Commands read from the repo's manifest — 1 month ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (12)

  • atlas

    Switch main-session persona to atlas (plan execution coordinator mode).

  • cancel-ralph

    Cancel the currently active Ralph Loop.

  • handoff

    Create a detailed context summary for continuing work in a new session.

  • hephaestus

    Switch main-session persona to hephaestus (deep autonomous execution mode).

  • plan

    Create a durable, reviewable implementation plan for a complex task.

  • prometheus

    Switch main-session persona to prometheus (planning discipline).

  • ralph-loop

    Start a self-referential development loop that runs until task completion.

  • selftest

    Validate Tier A behaviors inside a single Claude Code session.

  • sisyphus

    Switch main-session persona to sisyphus (orchestrator mode).

  • start-work

    Resume and execute tasks from an active implementation plan.

  • stop-continuation

    Global escape hatch to stop all automated continuation mechanisms.

  • ulw

    Enable ultrawork mode - a disciplined "continue until done" execution mode with parallel research and verification gates.

Subagents (9)

  • atlas

    Main-session execution coordinator persona for plan-driven delivery.

  • explore

    Contextual grep for codebases. Answers "Where is X?", "Which file has Y?", "Find the code that does Z". Fire multiple in parallel for broad searches.

  • hephaestus

    Main-session deep worker persona for autonomous implementation with strict verification.

  • librarian

    Specialized codebase understanding agent for multi-repository analysis, searching remote codebases, retrieving official documentation, and finding implementation examples using GitHub CLI, Context7, and Web Search.

  • metis

    Pre-planning consultant that analyzes requests to identify hidden intentions, ambiguities, and AI failure points.

  • momus

    Expert reviewer for evaluating work plans against rigorous clarity, verifiability, and completeness standards.

  • oracle

    Read-only consultation agent. High-IQ reasoning specialist for debugging hard problems and high-difficulty architecture design.

  • prometheus

    Main-session planning persona focused on high-quality markdown plans.

  • sisyphus

    Main-session orchestration persona focused on parallel exploration, execution, and verification.

Hooks (4)

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/scripts/hook_router.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/scripts/hook_router.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/scripts/hook_router.py

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/scripts/hook_router.py

View source on GitHub

Maintain claude-agent-kit?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted