Contributions to the project are always welcome. Please feel free to create pull requests, or report issues.
The primary way to contribute is by updating or adding more contexts in the lola-module/ directory and by extending or improving existing skills.
Note: The lola-module/ directory can package not only skills, but also commands, subagents, and MCP server components.
The lola-module/ directory contains the AI Context Module that packages all SecDevAI skills and contexts. It follows the Lola pattern for distributing AI skills across multiple assistants (Cursor, Claude Code, Gemini CLI, and more).
You can contribute by:
- Adding new security context files (e.g., language-specific patterns, framework-specific rules)
- Improving existing context files with better detection patterns or examples
- Adding new tool integration scripts
Each skill follows the Agent Skills pattern with a SKILL.md file that provides instructions for the AI assistant. You can contribute by:
- Improving existing skill instructions for better accuracy and coverage
- Adding new skills for additional security workflows
For guidance on adding new security context such as rules, tools, or languages, see the "Extending SecDevAI" section in docs/USAGE.md.
We encourage you to contribute your context extensions back to help the community! AI-generated code is absolutely welcome, however, just make sure to test your changes and provide clear descriptions in your PR. Every contribution, big or small, helps make SecDevAI better.
From the secdevai repository run:
uv run pytestInstall pre-commit hooks:
uv run pre-commit installFrom another repository run:
uvx --no-cache /path/to/secdevai