bmad-generate-project-context

To generate AI project context, this Skill performs repository discovery and creates a project-context.md file containing critical implementation rules, testing guidelines, and technology stacks. It uses step-based discovery to read project files and consolidate unobvious coding patterns into LLM-optimized documentation.

An LLM-optimized project context file is a concise markdown document containing unobvious implementation rules and edge case patterns. AI agents need it to implement code consistently without repeated repository discovery, ensuring they follow your specific technology choices and coding conventions.

Document repository rules by running this Skill to scan your project files and generate a tracked project-context.md. It captures technology stacks, testing rules, and deployment considerations, requiring user confirmation at key gates before appending finalized sections for AI agent consumption.

Generating project context does not require manual input for every convention. The Skill uses sequential, discovery-driven context generation to automatically read project files and extract critical implementation details, though it does require user confirmation before applying changes to the context file.

Yes, you can control the communication and output language. The Skill strictly adheres to configured language and document output settings during the repository discovery process, ensuring the generated project-context.md matches your preferred language for AI agent consumption.

The best way to capture unobvious implementation details is using a discovery-driven workflow that initializes tracked frontmatter and sequentially reads project files. This approach focuses on actionable rules and edge cases, producing lean output optimized for AI agents rather than exhaustive documentation.