source-driven-development

To ensure AI code follows official documentation patterns, you must enforce explicit source citations for every implementation decision and verify patterns against detected library versions to prevent deprecated APIs. This grounds code generation in verifiable sources.

Source-driven development is a pattern verification method that identifies and enforces official documentation-backed patterns for framework-specific code decisions. It ensures dependency on exact versioned patterns and embeds explicit citations in code comments.

To cite official docs in code comments, instruct the AI to fetch relevant official documentation for the detected stack, implement patterns exactly as documented, and include source citations directly in the output and maintain-facing code comments for fast verification.

Yes, documentation-backed coding verifies implemented patterns against detected framework or library versions. It prevents deprecated APIs by requiring exact versioned patterns from official sources before generating boilerplate or completing code reviews.

Version-aware implementation works for code reviews by checking framework-specific code decisions against official documentation. It validates exact versioned patterns and embeds explicit citations in the output to maintain correctness and avoid outdated implementations.

The primary limitation is its strict dependency on available official documentation. If a framework lacks comprehensive versioned docs, the skill cannot verify patterns, generate boilerplate, or enforce source citations for edge cases or undocumented features.