source-finder

Locate source code repositories for Python packages by analyzing PyPI metadata, project URLs, and code hosting platforms like GitHub, GitLab, and Bitbucket. Provides deterministic results with confidence levels.

⭐9stars🔀6forks📁View on GitHub🕐Updated Dec 23, 2025

When & Why to Use This Skill

The Source Finder Claude skill is a specialized utility designed to automatically locate the source code repositories of Python packages. By intelligently analyzing PyPI metadata and scanning platforms like GitHub, GitLab, and Bitbucket, it delivers deterministic repository URLs accompanied by confidence levels, significantly reducing the manual effort required for dependency research, security auditing, and open-source exploration.

Use Cases

  • Security Auditing: Efficiently locate the source code of third-party Python dependencies to perform security reviews and vulnerability assessments.
  • Open Source Contribution: Quickly find the official repository of a package to submit bug reports, feature requests, or pull requests.
  • SBOM Generation: Automate the collection of source code URLs for Software Bill of Materials (SBOM) and compliance documentation.
  • Dependency Verification: Confirm the authenticity and maintenance status of a library by verifying its linked project URLs and metadata.
namesource-finder
descriptionLocate source code repositories for Python packages by analyzing PyPI metadata, project URLs, and code hosting platforms like GitHub, GitLab, and Bitbucket. Provides deterministic results with confidence levels.
allowed-tools[Bash, WebSearch, WebFetch]

Source Finder

Locates source code repositories for Python packages with confidence scoring.

Usage

To find a source repository for a given package:

  1. Run the finder script, for example:
# Find repository
$ ./scripts/finder.py requests

# Output structure:
{
  "url": "https://github.com/psf/requests",
  "confidence": "high",
  "method": "pypi_metadata_project_urls.Source",
  "package_name": "requests"
}
  1. Parse the JSON output:
  • url: Repository URL (or null if not found)
  • confidence: high, medium, or low
  • method: How the URL was found
  • package_name: the package that was searched
  1. If confidence is low or url is null, use WebSearch: <package_name> python github repository

  2. Present results with confidence level clearly indicated

Output Format

As a result, provide structured output including:

  • Repository URL
  • Confidence level (high/medium/low)
  • Method used to find the repository
  • Additional context or warnings