Experimental laboratory for autonomous development and advanced debugging. Features in active development - use at your own risk.
You can install this plugin from any of these themed marketplaces. Choose one, add it as a marketplace, then install the plugin.
Choose your preferred installation method below
A marketplace is a collection of plugins. Every plugin gets an auto-generated marketplace JSON for individual installation, plus inclusion in category and themed collections. Add a marketplace once (step 1), then install any plugin from it (step 2).
One-time setup for access to all plugins
When to use: If you plan to install multiple plugins now or later
Step 1: Add the marketplace (one-time)
/plugin marketplace add https://claudepluginhub.com/marketplaces/all.jsonRun this once to access all plugins
Step 2: Install this plugin
/plugin install speclabs@allUse this plugin's auto-generated marketplace JSON for individual installation
When to use: If you only want to try this specific plugin
Step 1: Add this plugin's marketplace
/plugin marketplace add https://claudepluginhub.com/marketplaces/plugins/speclabs.jsonStep 2: Install the plugin
/plugin install speclabs@speclabsA Claude Code plugin marketplace for complete software development lifecycle management with spec-driven workflows, quality validation, and experimental autonomous features.
SpecSwarm provides two curated plugins that cover the complete software development lifecycle using Claude Code:
Complete lifecycle coverage from initial specification through implementation, bug fixing, modifications, refactoring, and sunset - with progressive enhancements for quality validation, tech stack consistency, and autonomous workflows.
Complete Software Development Toolkit
The unified, production-ready plugin for the complete development lifecycle. SpecSwarm consolidates spec-driven development, lifecycle workflows, quality validation, and performance monitoring into one comprehensive toolkit.
18 Commands | Production Ready | Phase 2.2 Complete
# New Feature Development
/specswarm:specify "Add user authentication with email/password"
/specswarm:plan
/specswarm:tasks
/specswarm:implement
# Bug Fixing
/specswarm:bugfix "Bug 915: Login fails with special characters"
# Code Quality
/specswarm:analyze-quality
/specswarm:refactor "Improve authentication module performance"
New Feature Workflows (8):
/specswarm:specify - Create detailed feature specification/specswarm:plan - Design implementation with tech stack validation/specswarm:tasks - Generate dependency-ordered task breakdown/specswarm:implement - Execute with comprehensive quality validation/specswarm:clarify - Ask targeted clarification questions/specswarm:checklist - Generate custom requirement checklists/specswarm:analyze - Cross-artifact consistency validation/specswarm:constitution - Create/update project governanceBug & Issue Management (2):
/specswarm:bugfix - Regression-test-first fixing with ultrathinking (Phase 2.2)/specswarm:hotfix - Emergency production issue responseCode Maintenance (2):
/specswarm:modify - Feature modification with impact analysis/specswarm:refactor - Metrics-driven quality improvementWorkflow Completion (1) ✨ NEW! Phase 2.2:
/specswarm:complete - Clean, guided feature/bugfix completion with merge to mainAnalysis & Utilities (5):
/specswarm:analyze-quality - Comprehensive codebase analysis/specswarm:impact - Standalone impact analysis/specswarm:suggest - AI-powered workflow recommendation/specswarm:workflow-metrics - Cross-workflow analytics/specswarm:deprecate - Phased feature sunsetExperimental Laboratory for Sprint/Project Orchestration & Advanced Debugging
⚠️ EXPERIMENTAL - Features in active development - Use at your own risk
The experimental wing of SpecSwarm, building toward autonomous sprint and project orchestration. The ultimate vision: coordinate multiple Claude Code agents to build features and complete projects with minimal human intervention.
Vision: "Give me a feature description Monday evening, wake up Tuesday morning with working, tested, production-ready code."
4 Commands | Phase 2 Complete (Oct 2025) | High Risk, High Reward
Impact: 50-67% faster feature implementation vs manual approach
Timeline: Phase 3 planned for Q1 2026
# Phase 2: Orchestrate Complete Feature (NEW!)
/speclabs:orchestrate-feature "Add user authentication" /path/to/project
# Phase 1b: Autonomous Task Execution
/speclabs:orchestrate features/001-fix/workflow.md /path/to/project
# Automated Validation
/speclabs:orchestrate-validate /path/to/project
# Systematic Multi-Bug Debugging
/speclabs:coordinate "navbar broken, sign-out fails, like button blank page"
Phase 2: Feature Orchestration (1) ✅ NEW!:
/speclabs:orchestrate-feature - Complete feature lifecycle from description to implementationPhase 1b: Task Orchestration (2) ✅:
/speclabs:orchestrate - Task workflow orchestration with intelligent retry logic/speclabs:orchestrate-validate - Validation suite (browser, terminal, visual)Advanced Debugging (1):
/speclabs:coordinate - Systematic multi-bug debugging with logging and orchestrationFuture Commands (Phase 3):
/speclabs:orchestrate-sprint - Complete sprint backlogs overnight (Phase 3)🎯 Use SpecSwarm v2.0.0 when:
🧪 Use SpecLabs v2.0.0 when:
| Feature | SpecSwarm v2.0.0 | SpecLabs v2.0.0 (Phase 2) | SpecLabs v3.0 (Future) |
|---|---|---|---|
| Feature Development | ✅ Manual with workflows | ✅ Fully automated | ✅ + Sprint-level |
| Bug Fixing | ✅ Regression-test-first | ❌ | ✅ Autonomous fixing |
| Modifications | ✅ Impact analysis | ❌ | ✅ Autonomous mods |
| Refactoring | ✅ Metrics-driven | ❌ | ✅ Autonomous refactor |
| Quality Validation | ✅ 0-100 point scoring | ✅ Playwright suite | ✅ Multi-layer validation |
| Tech Stack Enforcement | ✅ 95% drift prevention | ❌ | ✅ Autonomous enforcement |
| Bundle Size Monitoring | ✅ Performance budgets | ❌ | ✅ Autonomous optimization |
| Chain Bug Detection | ✅ Prevents cascades | ❌ | ✅ Auto-detection |
| Task Orchestration | ❌ | ✅ Full automation | ✅ Production-ready |
| Feature Orchestration | ❌ | ✅ Natural language → code | ✅ + Parallel features |
| Sprint Orchestration | ❌ | 🔄 Planned Phase 3 | ✅ Overnight sprints |
| Browser Automation | ❌ | ✅ Playwright | ✅ + Vision API |
| Advanced Debugging | ❌ | ✅ Systematic | ✅ Autonomous |
| Parallel Agent Execution | ❌ | ❌ | ✅ Multi-agent |
| Stability | ✅ Production-ready | ⚠️ Experimental | ✅ Production-ready |
| Commands | 17 | 4 | 5+ |
| Utilities | 8 | 5 components | 15+ |
| Timeline | Now | Oct 2025 | Q1 2026 |
Current (Phase 2): Autonomous feature development with SpecSwarm integration
Manual SpecSwarm Development:
/specswarm:specify → /specswarm:plan → /specswarm:tasks → /specswarm:implement
↓
/specswarm:analyze-quality (validate results)
Autonomous Feature Development (Phase 2) ✅ NEW!:
/speclabs:orchestrate-feature "Add user authentication" /path/to/project
↓
Automatically: SpecSwarm planning (specify → clarify → plan → tasks)
Automatically: Task conversion and implementation
Automatically: Validation and retry logic
Automatically: Bugfix if needed
↓
Deliver: Working, tested feature with comprehensive reporting
Future (Phase 3): Sprint-level autonomous development
Sprint-Level Orchestration (Phase 3):
/speclabs:orchestrate-sprint sprint-23-backlog.md /path/to/project
↓
Automatically: parse sprint backlog (identify features)
Automatically: dependency analysis (determine parallelization)
Automatically: execute features (uses orchestrate-feature for each)
Automatically: sprint coordination and tracking
↓
Deliver: Complete sprint ready for deployment
Debugging Integration:
SpecSwarm Single Bug → /specswarm:bugfix
SpecSwarm Multi-Bug (3+) → /speclabs:coordinate (systematic analysis)
# Add SpecSwarm marketplace to Claude Code
claude plugin marketplace add marty/specswarm
Or if using a local repository:
claude plugin marketplace add /path/to/specswarm
Install plugins from the marketplace:
# Install SpecSwarm plugin (production-ready complete toolkit)
claude plugin install specswarm
# Install SpecLabs plugin (experimental autonomous features)
claude plugin install speclabs
List available plugins:
claude plugin marketplace list
Spec-Driven Development is a structured approach to software development that prioritizes clear specifications before implementation:
This approach leads to:
Both plugins can be configured through files in /memory/:
/memory/quality-standards.md)# Quality Gates
min_test_coverage: 85
min_quality_score: 80
block_merge_on_failure: false
# Performance Budgets (SpecSwarm v2.0+)
enforce_budgets: true
max_bundle_size: 500 # KB per bundle
max_initial_load: 1000 # KB initial load
/memory/tech-stack.md)## Core Technologies
- TypeScript 5.x
- React Router v7
- PostgreSQL 17.x
## Approved Libraries
- Zod v4+ (validation)
- Drizzle ORM (database)
## Prohibited
- ❌ Redux (use React Router loaders/actions)
- ❌ Class components (use functional)
SpecSwarm automatically validates against this configuration at plan, task, and implementation phases.
If you were using older plugins, here's what happened:
All features from these plugins are now available in SpecSwarm v2.0.0.
Migration: Replace old plugin commands with SpecSwarm:
# Old
/speckit:specify → /specswarm:specify
/spectest:implement → /specswarm:implement
/speclab:bugfix → /specswarm:bugfix
Migration: Replace old plugin commands with SpecLabs:
# Old
/debug-coordinate:coordinate → /speclabs:coordinate
/project-orchestrator:orchestrate-test → /speclabs:orchestrate
/project-orchestrator:orchestrate-validate → /speclabs:orchestrate-validate
See DEPRECATED.md files in old plugin directories for full migration details.
specswarm/
├── .claude-plugin/
│ └── marketplace.json # Marketplace configuration
├── plugins/
│ ├── specswarm/ # SpecSwarm v2.0.0 (production)
│ │ ├── .claude-plugin/
│ │ │ └── plugin.json
│ │ ├── commands/ # 18 commands (NEW: complete)
│ │ │ ├── specify.md
│ │ │ ├── plan.md
│ │ │ ├── tasks.md
│ │ │ ├── implement.md
│ │ │ ├── analyze.md
│ │ │ ├── checklist.md
│ │ │ ├── clarify.md
│ │ │ ├── constitution.md
│ │ │ ├── bugfix.md
│ │ │ ├── hotfix.md
│ │ │ ├── modify.md
│ │ │ ├── refactor.md
│ │ │ ├── deprecate.md
│ │ │ ├── analyze-quality.md
│ │ │ ├── impact.md
│ │ │ ├── suggest.md
│ │ │ ├── workflow-metrics.md
│ │ │ └── complete.md # NEW! Phase 2.2
│ │ ├── templates/ # NEW! Phase 2.2
│ │ │ └── gitignore-diagnostic-files.txt
│ │ └── README.md
│ └── speclabs/ # SpecLabs v2.0.0 (experimental - Phase 2)
│ ├── .claude-plugin/
│ │ └── plugin.json
│ ├── commands/ # 4 commands
│ │ ├── orchestrate-feature.md # Phase 2
│ │ ├── orchestrate.md
│ │ ├── orchestrate-validate.md
│ │ └── coordinate.md
│ ├── lib/ # 17 utilities + components
│ │ ├── feature-orchestrator.sh
│ │ ├── task-converter.sh
│ │ ├── state-manager.sh
│ │ ├── decision-maker.sh
│ │ ├── prompt-refiner.sh
│ │ ├── metrics-tracker.sh
│ │ ├── vision-api.sh
│ │ ├── bundle-size-monitor.sh
│ │ ├── chain-bug-detector.sh
│ │ ├── orchestrator-detection.sh
│ │ ├── performance-budget-enforcer.sh
│ │ └── integration-patterns/ # NEW! Phase 2.1
│ └── README.md
├── README.md # This file
└── LICENSE # MIT License
Automated quality scoring across 6 dimensions (0-100 points):
Multi-Framework Support (11 test frameworks):
SpecSwarm is a consolidated plugin that builds upon SpecKit, which adapted GitHub's spec-kit for Claude Code.
Attribution Chain:
Original: GitHub spec-kit
Adapted: SpecKit plugin by Marty Bonacci (2025)
Enhanced: SpecSwarm v2.0.0 by Marty Bonacci & Claude Code (2025)
Key Consolidations (v2.0.0):
Experimental autonomous development and debugging with Phase 2 Feature Workflow Engine:
Created by: Marty Bonacci & Claude Code (2025)
Consolidated From:
Phase 2 Complete (October 16, 2025):
Based on Learnings:
docs/learnings/2025-10-14-orchestrator-missed-opportunity.mdThis project is licensed under the MIT License - see the LICENSE file for details.
Individual plugins may have their own attributions and licenses. See each plugin's README for details.
SpecSwarm Marketplace - Complete software development lifecycle in Claude Code. 🚀
Build it. Fix it. Maintain it. Analyze it. All in one place.
1.0.0