By sighup
Define feature specifications, break them into dependency-aware task graphs, and dispatch parallel agent teams to autonomously implement, review, test, and validate code against specs — all within Claude Code.
Bug fixer that investigates and fixes application bugs detected by tests. Fixes application code, never test code.
Coding worker that executes tasks using the 11-step protocol. Use to implement a specific task from the task board with atomic commits and proof artifacts.
Task planner that transforms specs into dependency-aware task graphs. Use when breaking down a specification into executable tasks with proper sequencing.
Read-only verification child that independently re-runs a task's proof commands and post-verification checks, returning a structured per-proof PASS/FAIL verdict. Use from cw-execute Steps 6 and 9 to gate task completion on observed results instead of the implementer's self-report.
Codebase researcher that performs deep exploration and produces structured research reports. Use before cw-spec to understand an unfamiliar or complex codebase and generate enriched context for specification writing.
Persistent agent team dispatcher with lead coordination. This skill should be used after cw-plan to execute tasks via a managed team (requires CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 and CLAUDE_CODE_TASK_LIST_ID).
Identifies independent tasks and spawns parallel agent workers. This skill should be used after cw-plan to execute multiple tasks concurrently.
Executes a single task from the task board using the 11-step implementation protocol. This skill should be used after cw-plan or cw-dispatch assigns a task, or when manually implementing a specific task by ID.
Internal subagent that generates Gherkin BDD scenarios from spec acceptance criteria. Produces one .feature file per demoable unit in the spec directory and optionally creates cw-testing task stubs on the task board. Called automatically by cw-spec.
Transforms a specification into a task graph with dependencies. This skill should be used after cw-spec to break a spec into executable tasks with proper sequencing before dispatching with cw-dispatch.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Claude Code plugin that unifies spec-driven development, autonomous task execution, and parallel agent dispatch into a single workflow. Takes a feature from idea to validated implementation using structured specifications, dependency-aware task graphs, and evidence-based verification.
# Add the marketplace
claude plugin marketplace add https://github.com/sighup/claude-workflow.git
# Install at project scope (shared with team via .claude/settings.json)
claude plugin install claude-workflow@claude-workflow --scope project
# Or install at user scope (personal, across all projects)
claude plugin install claude-workflow@claude-workflow --scope user
/plugin
# Navigate to Marketplaces tab → Add → paste the git URL
# Then go to Discover tab → select claude-workflow → choose scope
[/cw-research] → /cw-spec → [/cw-gherkin] → /cw-plan → /cw-dispatch → /cw-validate
Each step can also be run independently. /cw-execute handles single-task execution for manual or shell-scripted loops. /cw-review adds a code review gate and /cw-testing generates and runs E2E tests.
Use /cw-worktree to develop multiple features simultaneously. Each worktree gets its own feature branch and isolated task list (via .claude/settings.local.json). Tasks persist in ~/.claude/tasks/{worktree-name}/, enabling seamless resume across sessions. See examples/workflows.md for the full multi-terminal walkthrough.
| Skill | Purpose |
|---|---|
/cw-research | Explore codebase across 5 dimensions and produce a structured research report with /cw-spec meta-prompt |
/cw-spec | Generate structured specification with demoable units and proof artifacts |
/cw-plan | Transform spec into native task graph with dependencies and metadata |
/cw-execute | Execute one task using the 11-step protocol (orient → commit → clean exit) |
/cw-dispatch | Spawn parallel subagent workers for independent tasks (no setup required) |
/cw-dispatch-team | Persistent agent team with lead coordination for parallel task execution |
/cw-validate | Run 7 validation gates and produce a coverage matrix report |
/cw-review | Review implementation for bugs, security issues, and quality; creates fix tasks |
/cw-review-team | Concern-partitioned team review — each reviewer sees all files through a specialized lens (security, correctness, spec compliance) |
/cw-testing | E2E testing with auto-fix — generate tests from specs, execute, and fix failures |
cw-gherkin | Generate Gherkin BDD scenarios from spec acceptance criteria; called automatically by cw-spec |
/cw-worktree | Manage git worktrees for multi-feature parallel development |
jq is required by the worktree hooks and cw-status. The gh CLI is needed for PR creation.
Most skills work out of the box. /cw-dispatch-team uses Claude Code agent teams which requires two env vars:
// ~/.claude/settings.json
{
"env": {
"CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1"
}
}
// .claude/settings.json
{
"env": {
"CLAUDE_CODE_TASK_LIST_ID": "your-project-name"
}
}
Note: /cw-worktree create sets CLAUDE_CODE_TASK_LIST_ID automatically in .claude/settings.local.json — no manual configuration needed for worktree-based workflows.
/cw-dispatch (subagent workers) needs no setup and is the recommended default. /cw-plan will offer both options after task graph creation.
/cw-testing supports multiple backends. To use the playwright-bdd backend (Gherkin → Playwright, CI-friendly):
npm install --save-dev playwright-bdd @playwright/test
npx playwright install
herdr is a terminal-native agent multiplexer (Linux/macOS). When it is installed and running, /cw-worktree create will automatically open a Claude session in the new worktree's herdr pane instead of asking you to open a new terminal. /cw-worktree open <name> does the same retrospectively, focusing the existing pane if one already exists.
If herdr is not installed, not running, or you set CW_DISABLE_HERDR=1, every worktree command works exactly as before — the integration is silently skipped and the manual cd ... && claude instructions are printed instead. Nothing in the plugin requires herdr.
To give the in-pane Claude session full knowledge of herdr's CLI (workspaces, panes, wait output, etc.), install herdr's own skill into Claude Code after installing herdr:
Iterative optimization through automated evaluation. Evaluates artifacts against test cases, analyzes failures, generates targeted variants, and promotes winners.
npx claudepluginhub sighup/claude-workflow --plugin claude-workflowSpectacular skills/commands library for spec-anchored development and parallel task execution
Long-running agent harness with 5-layer memory architecture, GitHub integration, autonomous batch processing, Agent Teams with ATDD, 9 hooks (safety, quality gates, team coordination), and 6 Agent Skills
HarnessFlow — From idea to shipped product: high-quality engineering workflows for AI agents. Spec-anchored SDD, gated TDD, evidence-based routing, independent reviews, and formal closeout.
Requirements-driven development workflow with quality gates for practical feature implementation
GitHub Spec-Kit integration for constitution-based spec-driven development (7-phase workflow)
Skills-first specification-driven development framework with 7 agent skills for planning, implementation, review, and shipping. Natural language activation with intelligent agent orchestration. Includes /plan, /implement, /research commands plus managing-specifications, implementing-features, and reviewing-and-shipping skills.