lead-dev-os is a spec & context-aware SDD (Spec-Driven Development) framework for Claude Code development on small and large projects. It provides structured skills for product planning, spec writing, task scoping, and context-aware implementation.
The context of the project is stored under agents-context/ directory, it is domain specific and an agent is provided only the context it needs to perform its task.
Think about it as a lead developer's main tool for guiding a team of AI agents. A Lead Developer's primary responsibility is to ensure their team of engineers consistently moves in the right direction, making strategic decisions that benefit the entire project. By closely reviewing each engineer's progress before a feature is completed, the lead developer can identify potential issues early, address them proactively, and guarantee that the final product meets the highest standards of quality and reliability.
lead-dev-os is a Claude Code plugin, distributed via the captainme-ai plugin marketplace. Inside Claude Code, run:
/plugin marketplace add CaptainMe-AI/lead-dev-os
/plugin install lead-dev-os@captainme-ai
# Configure your project
/lead-dev-os:configure-project
The /lead-dev-os:configure-project skill will interactively set up your project — creating agents-context/ (with AGENTS.md, README.md, standards, and guides), lead-dev-os/specs/, and updating your CLAUDE.md.
To pick up new releases later, run /plugin marketplace update captainme-ai. See the Installation Guide for full details, including installing for development with --plugin-dir.
- Spec is the contract, code follows
- Requirements before implementation, always
- Agents execute against testable acceptance criteria
- Eliminates ambiguity-driven rework
- Inspirations from agent-os for spec-driven development, file organization and thourough prompting
- Composable — Load only what you need for the current task
- Self-referencing — Concepts link to related concepts
- Version-controlled — Track evolution of ideas over time
- AI-friendly — Agents load specific concepts as context before working
- Inspirations from agor for organizing the context
Full documentation is available at captainme-ai.github.io/lead-dev-os.
- Installation — Plugin setup, initialization, what gets created
- Workflow — The 4-step write → scope → implement → archive process
- Implementation — Autonomous, lead-in-the-loop, and hybrid execution modes
- Contributing — Testing and how to contribute
# Clone the repo (one-time)
git clone https://github.com/CaptainMe-AI/lead-dev-os.git ~/lead-dev-os
# Start Claude Code with the plugin
claude --plugin-dir ~/lead-dev-os/lead-dev-os
# Configure your project
/lead-dev-os:configure-project/lead-dev-os:configure-project → Set up agents-context, specs, standards, and CLAUDE.md
/lead-dev-os:plan-product → Define product mission, vision, users, tech stack
/lead-dev-os:define-standards → Establish coding style, architecture, testing conventions
/lead-dev-os:plan-roadmap → Create prioritized feature roadmap with phases
/lead-dev-os:create-or-update-concepts → Scan codebase to create/update concept files
/lead-dev-os:step1-write-spec → Interactive Q&A + formalize into spec with numbered requirements (FR-###)
/lead-dev-os:step2-scope-tasks → Break into task groups with context directives
/lead-dev-os:step3-implement-tasks → Context-aware implementation of task groups
/lead-dev-os:step4-archive-spec → Archive completed spec and block agent access
/lead-dev-os:create-pr → Open a GitHub PR for the current branch with a WHAT-focused description and emoji-prefixed title
See the Workflow Guide for the full process, and Implementation Modes for autonomous, lead-in-the-loop, and hybrid execution.
We are open to contributions — please open an issue or a pull request. See the Contributing Guide for testing instructions.
MIT License