Built for Claude Code

worca

Autonomous Software Development Pipeline

From requirements to production-ready code — orchestrated by AI agents, governed by safety gates, monitored in real-time.

Get Started View Docs
Preflight
Plan
Review
Coordinate
Implement
Test
Review
PR
Learn
0
Pipeline Stages
0
Specialized Agents
0
Event Types
0
Manual Steps

Nine stages. Zero gaps.

Every stage is purpose-built, from environment validation through post-run retrospective. Stages can be toggled, looped, and customized.

Preflight
Opus
Plan
Opus
Plan Review
optional
Opus
Coordinate
Sonnet
Implement
Sonnet
Test
Opus
Review
Opus
PR
Sonnet
Learn
optional

Seven specialists.
One mission.

Each agent has a focused role, curated tools, and strict guardrails. Opus handles reasoning-heavy work; Sonnet handles execution at speed.

P
Planner
Opus

Reads requirements, explores the codebase, and produces detailed implementation plans with file-level specificity.

requirements codebase analysis architecture
R
Plan Reviewer
Opus

Validates plans for feasibility, completeness, and correctness before execution begins. An optional quality gate.

validation feasibility optional
C
Coordinator
Opus

Decomposes plans into tracked tasks with dependencies, parallel execution groups, and clear acceptance criteria.

task decomposition dependencies parallelism
I
Implementer
Sonnet

Claims tasks and implements code using TDD. Writes failing tests first, then minimal code to pass. Parallel-ready.

TDD parallel code generation
T
Tester
Sonnet

Runs the full test suite, verifies coverage thresholds, and produces a proof status that gates the review stage.

test suites coverage proof status
G
Guardian
Opus

Reviews code quality, manages the PR lifecycle, and enforces human approval gates. The only agent allowed to commit.

code review PR management governance
L
Learner
Sonnet

Analyzes completed runs and produces ranked observations with actionable improvement suggestions for future pipelines.

retrospective improvements optional

Built for real-world
autonomous work.

Every feature exists because an autonomous pipeline needs it — not because it demos well.

Full Autonomous Pipeline

9 stages from preflight validation to post-run learning. Each stage loops, retries, and reports independently.

Governance & Safety

Guard hooks block dangerous operations. Test gates enforce coverage. Human approval checkpoints at every critical boundary.

Real-Time Dashboard

WebSocket-based monitoring with 52 event types. Global mode monitors all projects from one tab. Token costs, Kanban boards, and live logs.

Smart Coordination

Tasks with dependencies, parallel execution groups, and full traceability through the Beads task system.

Cost Control

Per-agent token tracking, circuit breakers for runaway failures, and configurable budget thresholds.

Resume & Recovery

Pause mid-stage, stop gracefully, resume from exactly where you left off. Full state persistence.

TDD-Driven

Implementers write failing tests first, then minimal code to pass. Guardian validates the full suite before PR.

Flexible Configuration

Agent prompt overlays, model selection per agent, stage toggling, loop limits, and per-project overrides.

Multi-Project Dashboard

Global mode monitors all registered projects from a single browser tab. Sidebar picker with live status dots and one-click project switching.

Parallel Pipelines

Run N pipelines concurrently, each isolated in its own git worktree. Per-pipeline pause, stop, and resume. Configurable cleanup policies.

Structured Feedback Loops

Four rework cycles — plan review, test failures, code review, and restart planning — each with configurable limits and bounded context.

Four ways to launch.

Start from a text prompt, a detailed spec, a GitHub issue, or run multiple pipelines in parallel.

from a prompt
# Describe what you need
python .claude/scripts/run_pipeline.py \
--prompt "Add user authentication
  with OAuth2 and JWT tokens"
from a spec file
# Use a detailed specification
python .claude/scripts/run_pipeline.py \
--spec spec.md \
--plan plan.md
from a GitHub issue
# Start from an existing issue
python .claude/scripts/run_pipeline.py \
--source gh:issue:42
parallel pipelines
# Run multiple pipelines concurrently
python .claude/scripts/run_multi.py \
--requests "Add auth" "Add search" \
--max-parallel 3

Up and running in minutes.

worca installs as a .claude/ folder. No global dependencies.

1
Start the dashboard

Launch the global dashboard to manage all your projects.

cd worca-cc/.claude/worca-ui && node server/index.js --global
2
Add your project

Click + in the sidebar to register and install worca in one step.

http://127.0.0.1:3400
3
Launch a pipeline

Describe your work and let the agents handle it.

run_pipeline.py --prompt "..."

Ship features while you sleep.

worca is open source and ready for your next project.

View on GitHub