Tooling discovery and analysis

[TomLucidor]

This all started when I am reading on this and it made me think https://repocheck.com/

• Source code https://github.com/dogweather/repo-health-check
• See also https://github.com/heff/repo-health?tab=readme-ov-file#all-projects
• For code complexity just using scc would be good https://github.com/boyter/scc

1. Whenever a tool is suggested, something like this should be used to examine if the codebase is dead or not
2. If it is alive and the codebase is not a mass, then using the code directly would be "best practice"
3. Assuming that the codebase is used, document common mistake (cus coding agents can get stumped multiple times)
4. If the codebase is dead, then capturing algos and migrating them into a move lively repo would be the best move
5. If it is alive BUT the codebase is a mess, then using a coding agent to refactor some of the code would be best
6. The PKM should record all these issues well such that the tooling-inspiration distinction can be made + "already read it" lists
7. Bonus: Avoid bad repos and proprietary tooling (including plugins) unless specified otherwise (e.g. "I want to use OpenAI")

Side note: some ideas are from personal experience with earlier versions of OpenHands