Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
By corca-ai
Orchestrate end-to-end development workflows using chained AI skills: gather context from URLs/codebase/searches, clarify requirements, draft/review plans, implement via domain agents, refactor code/plugins, run multi-reviewer quality gates (security/UX/domain), generate retrospectives, and automate GitHub issues/PRs/merges with human-in-loop and safety hooks.
npx claudepluginhub corca-ai/claude-plugins --plugin cwfUnified requirement clarification to prevent downstream implementation churn by resolving ambiguity early. Default: research-first with autonomous decision-making and persistent questioning. --light: direct iterative Q&A. Triggers: "cwf:clarify", "clarify this", "refine requirements"
Unified information acquisition that stabilizes context before reasoning: URL auto-detect (Google/Slack/Notion/GitHub/web), web search (Tavily/Exa), and local codebase exploration. Triggers: "cwf:gather", "gather <url>", "web search", "code search"
Auto-generate session or phase handoff documents so the next agent starts with context, constraints, and scope already loaded. Source: cwf-state.yaml and session artifacts. --phase mode generates phase-to-phase context transfer (HOW) separate from plan.md (WHAT). Triggers: "cwf:handoff", "cwf:handoff --phase", "handoff", "핸드오프", "다음 세션", "phase handoff"
Human-in-the-loop diff/chunk review to inject deliberate human judgment where automated review is insufficient, with resumable state, agreement-round kickoff, and rule propagation. Triggers: "cwf:hitl", "hitl", "interactive review", "human review"
Implementation orchestration from a plan to execute approved scope predictably without losing constraints. Decomposes steps into parallelizable work items, spawns domain-appropriate agents, and verifies completion against BDD success criteria. Triggers: "cwf:impl", "implement this plan"
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Production-ready Claude Code configuration with role-based workflows (PM→Lead→Designer→Dev→QA), safety hooks, 44 commands, 19 skills, 8 agents, 43 rules, 30 hook scripts across 19 events, auto-learning pipeline, hook profiles, and multi-language coding standards
AI-powered development workflow automation - Phase-based planning, implementation orchestration, preflight code quality checks with security scanning, ship-it workflow, and development principles generator for CLAUDE.md
Workflow skills and shared instructions for coding agents.
Verification-first engineering toolkit for Claude Code. 15 skills across a 5-phase spine (Investigate → Design → Implement → Verify → Ship), 8 specialist agents, an interactive setup wizard. Every skill has rationalizations + evidence requirements. Built for senior ICs and tech leads.
Development workflow automation plugin: specify → open → execute pipeline with parallel research agents, hook-based guards, and PR state management
Framework for AI-powered multi-step workflows with quality gates
Standalone intentful behavior evaluation CLI and bundled skill.
Modifies files
Hook triggers on file write and edit operations
Modifies files
Hook triggers on file write and edit operations
Share bugs, ideas, or general feedback.
Disclaimer (SoT)
README.ko.mdis the single source of truth for CWF user-facing policy. If docs and implementation diverge, fix the plugin to match docs and report the mismatch via issue/PR.
A Claude Code plugin that turns structured development sessions into a repeatable workflow — from gathering context through retrospective analysis. Maintained by Corca for AI-Native Product Teams.
# Add the marketplace
claude plugin marketplace add https://github.com/corca-ai/claude-plugins.git
# Install CWF
claude plugin install cwf@corca-plugins
# Restart Claude Code for hooks to take effect
After restart, run one-time bootstrap in Claude Code / Codex CLI:
cwf:setup
cwf:setup standardizes first-run behavior by handling:
.cwf-config.yaml, .cwf-config.local.yaml)shellcheck, jq, gh, node, python3, lychee, markdownlint-cli2)For detailed flags, see setup.
If you also use Codex CLI, running only cwf:setup is enough to get guided defaults. Codex integration now follows the active plugin scope by default (local > project > user). Use the commands below when you want to re-apply Codex integration only:
cwf:setup --codex
cwf:setup --codex-wrapper
In Claude Code / Codex CLI, start with a plain prompt:
I need to solve <problem>. Please use CWF and drive the workflow.
The agent can invoke cwf:run and chain gather → clarify → plan → review(plan) → impl → review(code) → refactor → retro → ship. If automated review is not enough, switch to cwf:hitl to document key decision points and user concerns first, then continue chunk-by-chunk review.
You do not need to memorize every skill. The sections below explain why each skill exists and when to use it.
claude plugin marketplace update corca-plugins
claude plugin update cwf@corca-plugins
Or from inside Claude Code / Codex CLI:
cwf:update # Check + auto-update if newer version exists
cwf:update --check # Version check only
As of v3.0.0, legacy standalone plugins have been removed from the marketplace. If pre-v3.0 standalone plugins are installed, uninstall them and install cwf instead.
cwf) that integrates context gathering, requirement clarification, planning, implementation, review, retrospective, handoff, and shipping.$200 Max plan users).