Execute Complex MissionsWith One /coord Command
Getting Started with AGENT-11
New to AGENT-11? Follow this step-by-step guide to go from installation to executing your first mission in under 5 minutes.
Installation takes less than 1 minute. Your first mission can be running in 5 minutes.
Install AGENT-11 NowThe Solo Founder's Dilemma
Building solo means wearing every hat. But quality, speed, and sanity shouldn't be the price you pay for independence.
Execute Complex Missions with /coord
From BUILD to MVP to DEPLOY - watch how project-local agents collaborate to complete missions faster than traditional development teams.
THE STRATEGIST
Elite product strategy specialist. Rapid MVP definition, user stories in INVEST format, laser focus on shipping.
# User Authentication Stories
## Epic: Secure User Access
**Value**: Enable users to securely access personalized features
### User Stories:
1. **As a** new user **I want to** create an account **so that** I can save my preferences
- Acceptance: Email validation, password strength check
- Effort: 3 points
- Priority: P0 (Must Have)
2. **As a** returning user **I want to** log in quickly **so that** I can access my dashboard
- Acceptance: Remember me option, error handling
- Effort: 2 points
- Priority: P0 (Must Have)
## Definition of Done:
✅ Secure password hashing (bcrypt, 12+ rounds)
✅ Email verification flow implemented
✅ Rate limiting protection (5 attempts/15min)
✅ Session management with JWT
✅ Password reset functionalityReady to Deploy Your AI Development Team?
This is the squad I build my own products with, open-sourced so you can inspect it, verify it, and run it yourself. Free and MIT-licensed.
Follow the build
AGENT-11 is developed in the open. Get a note when a new release lands, with what changed and why.
The Ultimate Dogfooding Proof
3-Week Development Timeline
Format Standardization
🔧 Key Activities
- ✓Standardized agent metadata format across all specialists
- ✓Established file-based agent storage in .claude/agents/ directory
- ✓Prevented scope creep through explicit task boundaries
- ✓Orchestrated sequential specialist pipeline approach
🎯 Outcome Achieved
💡 Key Insight
Choose Your Development Strategy
Production Pipeline Pattern
Building production systems, deploying critical infrastructure
Specialist Pipeline
THE COORDINATOR
Project orchestration
🎯 Expected Outcome
Production-ready system with comprehensive documentation and testing
💡 Real Example
Used to build AGENT-11's own deployment system, with the whole build open to read
Ready to Deploy Your Specialist Squad?
These proven workflows are ready to use. The same patterns that built AGENT-11 itself.
“AGENT-11 built its own deployment system. This isn't theoretical: the whole build is open, so you can read the code and the decisions for yourself.”
Join the Community
From Scratch to Production-Ready
Stop building everything from scratch. Your specialists deliver production-ready results instead of the half-finished output of a single person juggling every role.
In my own builds, the squad is orders of magnitude faster than working solo. I don't publish benchmark numbers; the code is open so you can judge for yourself.
Real Example: AGENT-11 built AGENT-11
"AGENT-11's own deployment system was built by AGENT-11. The whole build is open: read the code, the case study, and the decisions for yourself."
Built for Developers, By Developers
Every technical decision is in the open. Plain files in your repo, run inside Claude Code, nothing hidden behind a service.
How the squad is put together
AGENT-11 is a set of Markdown agent definitions installed into your project and run inside Claude Code. No runtime to host, no service to run.
File-based agents
Each specialist is a Markdown file in .claude/agents/. You can read, edit, and version-control every one.
Coordinator orchestration
The /coord command runs a mission: the coordinator delegates to specialists and gates each phase on evidence.
Persistent context
A small set of tracking files (project-plan.md, agent-context.md, evidence-repository.md) carries state across sessions and handoffs.
Read-only quality gates
The criteria that judge the work are unwritable by the agents doing it, so a passing gate means the work was done.
# Installed into your project .claude/ agents/ # 11 specialist definitions (Markdown) commands/ # /coord and the other slash commands settings.json # tool permissions + read-only gate rules # Run a mission /coord build requirements.md /coord continue # phase-gated: advances only on evidence
Built On
Questions people actually ask
Short answers, and where to check them for yourself.
- What is AGENT-11?
- AGENT-11 is an open-source multi-agent development framework for Claude Code. One install command deploys 11 specialist agents — coordinator, strategist, architect, developer, designer, tester, documenter, operator, analyst, marketer and support — into your project's .claude/ directory. You then run work as missions through the /coord command.
- How do I install AGENT-11?
- Run this from your project directory: bash <(curl -fsSL https://raw.githubusercontent.com/TheWayWithin/agent-11/main/project/deployment/scripts/secure-install.sh). Every install deploys all 11 specialists, and the coordinator pulls in whichever the mission needs. There is no squad size to pick.
- How much does AGENT-11 cost?
- AGENT-11 is free and open source (MIT). There are no licence fees and no per-mission charges. Your only cost is your own Claude Code / Anthropic API usage, which depends on how much you run it.
- Where does my code go when I run a mission?
- AGENT-11 is a set of files installed locally in your project's .claude/ directory; the agents and your code stay in your repo. Like any Claude Code session, the code and context you work on are processed by Claude (Anthropic). There is no AGENT-11 server, and nothing is sent to us.
- Which version is current?
- v6.2.0, released 20 June 2026. It added read-only quality gates, default-fail verification, the ratchet optimiser and a phase-gated meta-loop in /coord continue. The full history is on the changelog page and in the repository CHANGELOG.
- Who builds AGENT-11, and how is it funded?
- Jamie Watters builds and maintains it as a solo founder, in the open. The source is on GitHub under the MIT licence, so you can read every line before you run it. Nothing on this site is for sale; support is voluntary, through GitHub stars or Buy Me a Coffee.
How this project works
AGENT-11 is built and maintained by Jamie Watters, a solo founder, with Claude (Anthropic) doing much of the coding under his review — the framework was used to build itself, and that method is the subject of the project rather than a secret behind it. It is free under the MIT licence. Nothing on this site is for sale, there are no sponsors and no affiliate links, and support is voluntary. Every claim here can be checked against the source on GitHub. Site last updated .
Ready to Build With a Full Squad?
Trade your solo struggle for 11 specialists working alongside you. Your personal dev team is minutes away.
From Zero to Deployed in 3 Steps
Deploy Your Squad
One command installs all 11 specialist agents into your project.
Start Building
Run a mission with deterministic dispatch — no NLP guessing about intent. 13 missions across greenfield, surgical, and maintenance modes.
Deploy & Scale
Production-ready code with tests, documentation, and deployment guides
Your Full Squad
Every install deploys all 11 specialists, ready for your codebase from the first mission
100% Free Forever
Every agent. Every mission. Every feature. No trials. No limits.
100% Free & Open Source
Fork it. Tweak it. Make it yours. AGENT-11 is fully open source under MIT license—copy the code, customize the agents, adapt it to your workflow, or just benefit from ongoing development. No strings attached.
- All 11 specialized agents
- Complete mission framework
- Unlimited projects
- Full documentation access
- Community support
Support & Connect
Solo dev building this for founders like you. Every coffee fuels more features. Want to chat? I'm always happy to connect.
- • Maintain servers & docs
- • Build new agents & missions
- • Provide community support
- • Keep everything free forever
Connect with Jamie:
Stop wearing 11 hats. Get 11 specialists.
Quick setup • No credit card required • Open source forever