building-github-index-v2
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The scripts
github_index.pyandpk_index.pyperform network requests to GitHub APIs (api.github.com) and content delivery domains (raw.githubusercontent.com). These operations are used to fetch repository trees, file contents, and archive tarballs. As these target well-known and trusted infrastructure for the purpose of repository indexing, they are considered safe. - [COMMAND_EXECUTION]: The skill provides Python scripts designed to be executed via the command line to generate markdown indexes. These scripts use standard libraries for file system operations and network requests. The instructions also include example
curlcommands for manual API interaction. - [CREDENTIALS_UNSAFE]: The skill requests GitHub Personal Access Tokens (PAT) via CLI arguments (
--token) or environment variables (GITHUB_TOKEN,GITHUB_PAT). This is standard practice for tools interacting with the GitHub API and does not involve hardcoded secrets or unsafe storage of credentials. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from GitHub repositories (Markdown headings, code symbols, YAML frontmatter). While this creates a surface where a malicious repository could include injection attempts in metadata, the skill implements basic sanitization (e.g., removing HTML tags from frontmatter) and structural formatting (Markdown tables) that reduces the risk of an agent misinterpreting index data as instructions. The risk is assessed as low given the intended use case.
Audit Metadata