claude

agentkit

One coordinator, role profiles with per-role model pinning, deterministic zero-token checkpoints, launch preflight, and memory that stays with the coordinator instead of following every disposable worker.

0 on GitHubby Abdülkadirv0.1.1MITupdated 2 months agoworkflow
/

Plugin agentkit from marketplace agentic-kit in Akadirr1/agentic-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 (2)

  • agentkit-launch

    How to start a CLI agent as a worker so it reaches an input-accepting state instead of stopping on a prompt nobody answers. Use when a supervised worker fails at agent_readiness, when an agent pane looks hung, when choosing between a supervised worker and a one-shot call, or when deciding which perm

  • agentkit

    Set up or operate the agentkit coordination layer in a project — scaffolding agents/, choosing which layer a piece of work belongs to, keeping one coordinator with disposable workers, and checkpointing state without a model. Use when a project needs multi-agent coordination, when deciding whether to

Slash commands (4)

  • /agentkit:doctor

    Verify the agentkit layout, roles, CLIs, hooks, and knowledge stores

  • /agentkit:init

    Scaffold the agentkit coordination layer here, then fill it from this repository

  • /agentkit:preflight

    Prove every role in roles.toml actually launches unblocked

  • /agentkit:update

    Refresh the kit-managed files in this project and report what changed

Subagents (6)

  • analyst

    Diagnose failures from logs, traces, timing, and resource data without changing code. Produce hypotheses ranked by the evidence that would settle them.

  • architect

    Analyse cross-component structure, failure boundaries, and data flow before implementation, and propose the smallest change that respects existing contracts.

  • reviewer

    Independently review a diff together with its structural neighbourhood, covering correctness, regressions, contract breakage in unchanged callers, and acceptance-criterion coverage.

  • safety

    Independent hazard and fail-safe review for changes that can cause harm, data loss, or irreversible action.

  • scribe

    Exclusive writer for accepted project documentation. Writes canonical state directly, gated on evidence and a fixed scope limit rather than on a human approving each delta.

  • test-verifier

    Independently execute and assess checks after implementation, distinguishing what was verified from what merely ran.

Hooks (3)

  • SessionStart

    [ -f agents/checkpoint.sh ] || exit 0; grep -q checkpoint.sh .claude/settings.json 2>/dev/null && exit 0; bash agents/checkpoint.sh resume

  • Stop

    [ -f agents/checkpoint.sh ] || exit 0; grep -q checkpoint.sh .claude/settings.json 2>/dev/null && exit 0; bash agents/checkpoint.sh save

  • SessionEnd

    [ -f agents/checkpoint.sh ] || exit 0; grep -q checkpoint.sh .claude/settings.json 2>/dev/null && exit 0; bash agents/checkpoint.sh close

View source on GitHub

Similar plugins

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development

9 skills · 1 command · 8 agents

wshobson/agents 42,461

beads

claudecodex

AI-supervised issue tracker for coding workflows

1 skill · 1 agent · hooks

gastownhall/beads 28,698

superset

claude

Use Superset to spawn agents, create workspaces, manage tasks, and schedule automations from Claude Code.

14 skills · 1 command

superset-sh/superset 15,548

MiniMax AI skills library for frontend, fullstack, Android, iOS, shader, GIF sticker, document, presentation, spreadsheet, and multimodal media workflows

18 skills

MiniMax-AI/skills 13,828

worktrunk

claudecodex

Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows. This plugin provides configuration guidance (LLM commit messages, project hooks, worktree paths), automatic activity tracking (🤖/💬 indicators in `wt list` showing active Claude sessions), and a `/wt-diff` pane showing each worktree's changes.

2 skills · hooks

max-sixty/worktrunk 12,481

Maintain agentkit?

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