Claude Code and Codex plugins their authors filed under "workflow". All categories.
Verification-gated autonomous loop engineering: Claude works act→verify→repeat until a verify command passes, bounded by iteration caps and no-progress detection.
1 skill · 4 commands · 2 agents · hooks
Idea to shipped software as a sequence of reviewable artifacts: refine the spec, architect it, break it into tasks, then build under a working agreement with real gates. Includes the phase-boundary ritual, the issue-review loop, and a catalogue of the ways a green test suite lies to you.
8 skills
Switch Claude Code model using a live proxy model list plus optional local metadata seed.
7 commands · 1 MCP server
網頁版：需求、開卡、Bug 通報、PR feedback 等 7 個不需要本機 checkout 的流程。
8 skills
1 command (/patterns) + 6 executor agents + 8 pattern templates (agent-monitoring, code-deep-research, collaborative-dev, committee-review, coordinator-analysts, new-project, research-review-committee, scratch-protocol). Supports list / instantiate / --patch / --sync-check modes. Instances carry versioned generated-from fields and are registered for drift tracking.
1 skill · 1 command · 6 agents
Run an AI review-and-fix loop until it converges: against a reviewer bot on a pull request, or against a review command on your machine.
1 skill · 7 commands
Revisor operacional anti-burocracia: skills /doctail:review, simplify, audit, diff, style; subagents doc-auditor e doc-refactorer; hooks de guarda e alerta pós-edição.
5 skills · 2 agents · hooks
Slash-command wrapper around the agent-handoff CLI — /agent-handoff:init, /agent-handoff:push, /agent-handoff:pull, /agent-handoff:diff, /agent-handoff:status, /agent-handoff:doctor, /agent-handoff:bootstrap, /agent-handoff:deps.
8 commands
Incremental implementation review, PR finalization, reviewer-team, and Slack review-request workflows for Xedoc, Claude Code, and Codex.
7 skills · hooks
A task router: detect the task type (research · design · implement · debug · review · optimize) and run a pre-curated, model-tiered workflow — a shared prelude that unblocks a parallel fan-out (Sonnet builds, Haiku scouts), then a convergence. Opus orchestrates by default, swappable to Fable. Ships the /orchestrate router, an /orchestrate-workflows companion, pre-tiered subagents, a SessionStart doctrine, and layered repo+user config.
2 skills · 5 agents · hooks
Spec 기반 기능 워크플로우 관리 플러그인 - 스펙 생성, GitHub 게시, 기능 구현 계속, 리뷰 이슈 수정
4 skills · 4 commands
Commander session skills: roll-call of named worker sessions against a roster, dispatch of self-contained handoffs with a reply contract and a ledger, recommendations to open parallel sessions when work queues, and questions to the owner through Atlas with answers and agent messages delivered into every session
7 skills · 1 agent
One manager session runs spec-gated Orca workers, one per PR, with independent review and confirmed cleanup.
1 skill · 1 command · 5 agents · hooks
A four-role dev crew (Planner, parallel Coders and Testers, and a fan-out Reviewer) that builds and reviews features — or stands up a whole new codebase — in the requestor's own style, run as a Claude Code agent team whose teammates coordinate through the shared task list and mailbox. Auto-engages for feature and new-project work; also exposes /4man:code-review. Requires the experimental agent-teams feature (CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS).
1 skill · 1 command · 6 agents
Swift and Xcode instantiations of the vvkit disciplines. Install per project, with --scope project
6 skills · 1 agent · hooks
The engineering partner's playbook for executing an Approved Plan: scaffold Notion Plan → GitHub, the issue contract, Definition of Done, the governed-repo engineering flow, the pre-PR review gate, the review-response contract, the pre-DoR self-review gate, and Notion MCP editing hygiene.
8 skills
Fluxo de produto da mK: ficha-do-projeto levanta as fontes e gera a Ficha do Projeto (etapa 3), mapa-do-projeto publica o canvas Conceito, Capacidades, Roadmap e Hipóteses, projeto-issues quebra o projeto aprovado em issues no Linear mK, linha-do-tempo monta as fases em barras com o bloco "Depois", esboco explora formatos novos e registra o escolhido. Motor copiado do VANTA, camada da mK.
8 skills
/work:* commands for git worktrees with automatic session context.
6 commands · hooks
Product-Owner-led team workflow: a Product Owner orchestrates specialist subagents (architect, designer, engineer, qa, reviewer, docs) that collaborate through GitHub issues and pull requests.
7 skills
Team engineering rules (AGENTS.md), the plan-conventions, review-copilot-comments, sync-agent-rules and sync-memory-upstream skills, and a PreToolUse credential guard.
6 skills · hooks
/harness:run · /harness:uiux · /harness:resume 슬래시 커맨드 + 스킬 번들 + 멀티세션 하드닝(HARNESS-MSC-001) 포함
3 skills · 3 commands · hooks
Six process skills that enforce read-only discovery, spec-before-plan, plan-before-code, TDD, root-cause debugging, and evidence-before-claim. Stack-agnostic: per-project commands come from a .claude/subpowers.md contract.
6 skills · hooks
Token-lean agentic dev flow on Orca: fresh workers, gated tasks, independent audit, and per-model token measurement.
7 skills
Spec-driven delivery protocol: role agents (planner/builder/tester/reviewer/release), a gsd workflow skill, and a verification gate that blocks commits until your .gsd/workflow.json checks pass.
1 skill · 5 agents · hooks
Carries verified work context between developers' agent fleets, and decides on your behalf from a policy you wrote.
6 commands · hooks
Distills repo facts, session telemetry, and skill-health signals into durable, evidence-backed agent memory.
3 skills · 1 command · 1 agent · 1 MCP server · hooks
Analyze client RFP/BOQ packages, assess staffing gaps against company capabilities, and generate complete commercial proposals with pricing estimates
2 skills · 5 commands
Spec-driven design system: visual brainstorming with forge, structured plans with SDD defaults, subagent/agent-team execution with TDD and worktree hygiene
3 skills · 3 commands · hooks
The complete Spec-Driven Development workflow: spec-creator, implementation-planner, implementer, and plan-verifier agents plus the run-plan orchestrator, manual workflow-retro, and engineering-insights capture — wired to shared knowledge skills, a research agent, and an architecture-review gate via plugin dependencies.
3 skills · 4 agents
git blame for AI decisions: capture why agents chose what they chose, anchor it to code, and expose it to future sessions via an MCP server. Includes a staleness detector, decision graph, confidence-weighted rollups, and an EU AI Act JSON-LD export with HMAC / Ed25519 signing.
5 commands · 1 MCP server · hooks