search-first

To find existing libraries before writing new code, search across language registries like npm and PyPI, MCP servers, GitHub, and web sources to compare candidates and assess fit for your requested capability.

Evaluating tooling for code reuse involves scoring candidates on functionality, maintainability, documentation, license, and dependencies to return structured guidance on whether to adopt, extend, wrap, or build the solution.

Yes, you can search MCP servers, GitHub, Claude Code skills, language registries, and web sources to identify existing solutions and patterns, ensuring you locate at least three candidates before writing custom code.

Deciding whether to wrap or build a library requires comparing existing candidates against your feature gaps and scoring them on maintainability and dependencies, yielding actionable adoption, extension, wrapping, or building guidance.

Researching existing solutions before coding prevents reinventing the wheel by systematically locating existing tools and patterns, ensuring you identify at least three candidates and one gap for a given feature request.

Yes, the research workflow supports language registries including npm and PyPI, alongside MCP servers and GitHub, to locate existing tools and assess their fit before writing new code.