Workflow plugins

Claude-led team harness: scaffold multi-agent config (Codex, Cursor) from a single source of truth (AGENTS.md).

25 skills · 21 commands

The full agent OS: 18 skills, 26 agents, lifecycle hooks + always-on sensors, rules, RL telemetry, cross-family verification, prediction ledger

18 skills · 26 agents · hooks

Full-stack software consultancy pipeline — 8 specialized AI agents covering the complete SDLC from lead qualification through architecture, delivery planning, implementation, testing, release, and ongoing operations. Built for professional consulting engagements.

25 skills · 9 commands · 10 agents

Scaffold docs, drive a TDD roadmap loop, run adversarial reviews that auto-file tech-debt, design by judge panel, and ship via gated Conventional-Commit releases.

9 skills · 13 commands · 16 agents · 1 MCP server · hooks

코딩 에이전트용 '닫힌 루프(closed-loop)' 워크플로 40종의 한국어판. 테스트/CI/리뷰/배포/보안/문서/마이그레이션 등 목표 달성까지 자동 반복하는 루프를 한글 설명과 함께 제공합니다.

40 skills

30 skills, 5 agents, and 5 hooks for full-stack development with Next.js 14 + Python FastAPI. Covers brainstorming, TDD, code review, systematic debugging, security audit, refactoring, and the gbrain/CodeMAX shared-brain workflow.

32 skills · 3 commands · hooks

A secretary to coding, not the coder: skills that keep a project's record, roadmap, docs and tooling honest, with one writer per record file.

24 skills · 2 commands · 6 agents · hooks

Claude Code and Codex coordination bus, durable shared ledger, host-attributed collision warnings, authenticated relay replay, skill workflows, and opt-in hard anti-clobber enforcement. Warn-only by default.

3 skills · 19 commands · 10 agents · hooks

A startup crew for Claude Code: pitch, blueprint, sprint, break, audit, pr. 13 agents, 20 commands.

19 skills · 13 agents

Base toolset for personal development workflow and utilities

29 skills · 1 MCP server · hooks

Plan -> implement -> review -> PR -> converge autopilot, cross-model (Claude<->Codex) review, product/design/UX critique, and evidence-driven competitive research. Composable slash commands that scale ceremony to the task and never auto-accept a reviewer.

23 commands

eSoul's own skills, including the esoul-workflow entry point that sizes a task and routes it to the right skill.

15 skills · 7 agents · hooks

Cairn — marks the trail and remembers the path. Batteries-included: installing it pulls in GSD (a plugin dependency) and bootstraps the beads issue tracker (bd), then a single /cairn:init wires a project end to end. Wires the GSD planning workflow (/gsd:*) to beads (bd) so phases create/claim/close tracked work, optionally mirrors bd issues two-way (hub-and-spoke) to GitHub Issues, GitLab, Jira, Asana, and Azure Boards, and optionally makes the context-mode knowledge base intent-aware (scopes compressed memory to the active issue + phase). The integration conventions activate only in repos that use both GSD and beads.

3 skills · 18 commands · hooks

Workflow commands, specialized subagents, safety hooks, and reusable skills for feature development.

16 skills · 5 agents

Ten agent roles, four approval gates, GitHub as pipeline state and OpenSpec as spec content. Ships the role commands, the protected-branch hook, and the reusable GitHub Actions workflows that drive them.

6 skills · 13 commands · hooks

Multi-agent kanban workflow for Claude Code: a JSON task board, coordination state, enforcement hooks, and skills for planning, parallel execution, handoffs, and close-out.

16 skills · 2 agents · hooks

State-file memory, lookup-driven ticks, blocking hooks, a stop gate, an escalation door, a regression diff with a baseline, and a model-checking driver — everything that makes Claude finish work end to end.

10 skills · 5 commands · 3 agents · hooks

Research → Position → Create → Gate → Measure for marketing work. A MARKETING_MAP.md context file, tiered session boot, and hooks (session-start, prompt-router, branch-protect, block-dangerous-commands). Skills, gates and agents land in later 0.1 work.

15 skills · 3 agents · hooks

Workflow completo de desenvolvimento do Murilo (AusTV Dev), fundido com as skills do superpowers vendorizadas. Auto-inicia em toda sessao.

16 skills · 1 agent · hooks

Repository workflow driven by a project profile, from the task's plan to promotion, with a fresh session per phase, a small change in one. GitHub with gh; no tracker.

5 skills · 12 commands · 1 agent

Planning, diagnosis, GitHub and review workflows

17 skills

Atomic Work Units, phase execution contracts, verification evidence, and client-confirmed closeout.

3 skills · 14 commands

End-to-end development pipeline with evidence-rich finalize and ship.

15 skills · hooks

BMAD Method as native Claude Cowork teams — 9 specialized AI agents (Analyst, PM, UX, Architect, Scrum Master, Developer, QA, Quick Dev, Tech Writer) that collaborate autonomously through phased workflows.

3 skills · 3 commands · 9 agents

Complete software development methodology (14 skills): brainstorming, TDD, systematic debugging, writing/executing plans, subagent-driven development, code review, git worktrees

14 skills

Structured AI development framework. Adaptive complexity, epistemic rigor, total ownership, 9 specialized agents, and pre-push quality gates.

1 skill · 2 commands · 9 agents · hooks

Wayfind uncertain intent into an Atlas, chart each expedition as a GitHub DAG, then pre-flight, execute and prove it to done-clean.

13 skills

AI-SDD v3 方法论插件 -- 提供完整的规范驱动开发工作流

13 skills

Cross-model shipping pipeline on six pinned models: Astra plans, Fable, Grok and a fresh Astra attack the plan, Luna, Grok and Sonnet implement in parallel worktrees, Opus gates each task, and Fable, Astra and Grok all attack the branch before the PR.

1 skill · 1 command · 11 agents

Opinionated GitHub workflow: worktree + draft PR flow, self-documenting issues, and an independent PR reviewer bot.

10 skills · 2 agents · hooks