| name | GitHub Repository Analysis |
| description | Analyzes a github repository to extract key features, innovation score, and categorization. |
| category | Code Analysis |
| tags | ["github","code","repository"] |
| version | 1.0.0 |
GitHub Repository Analysis Skill
Objective
To extract high-quality, normalized information from a GitHub repository README and source structure to classify the tool effectively into the Borg Atlas.
Triggers
This skill activates whenever a github.com URL is passed into the pipeline.
Execution Strategy
- Fetch & Clean: Retrieve the
README.md. Strip out badges, contributor lists, and table of contents to focus on the core descriptions.
- Feature Extraction: Do not just list "Open Source". Extract the core architectural features (e.g. "Uses AST parsing for deterministic type generation", "Implements zero-copy memory buffers").
- Innovation Scoring:
- Baseline = 5.
- If it implements a completely novel algorithm not seen in other repos -> +2
- If it is a wrapper around an existing tool without new features -> -2
Prompt Modifiers for LLM
- Focus on: stars, language, last commit date, README quality
- Extract: primary use case, key dependencies, API surface
- Prioritize: how this differs from alternatives in the same space
Output Schema
Ensure output complies with the standard atlas entry JSON structure.