Zum Inhalt springen
Kitbay
Skill

dependency-doctor

Apache-2.0Aktualisiert am 11. Nov. 2026144.402GitHub-SterneEnthält Skripte

A skill by Matt Van Horn in the awesome-llm-apps repository that checks one dependency file the user names for surface-level mistakes.

Was der Skill vermittelt

A skill by Matt Van Horn in the awesome-llm-apps repository that checks one dependency file the user names for surface-level mistakes. A standard-library-only Python script reads requirements.txt, pyproject.toml (PEP 621 or Poetry tables) or package.json and reports, as JSON, packages that shadow Python standard-library names, backports that should not be installed, unconstrained versions, repeated entries and conflicting pins, each with a suggested fix. It runs offline; only with the user's explicit approval does an online mode ask pypi.org whether pinned releases were yanked. It does not check for vulnerabilities or resolve dependencies, and it edits the file only after approval.

Funktionen

  • Check requirements.txt, pyproject.toml or package.json for problems
  • Flag unconstrained, repeated or conflicting dependencies
  • Flag packages that shadow Python standard-library names
  • Check pinned PyPI releases for yanks, with approval