binlee-source-library

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local utility scripts (scripts/search-corpus.mjs and scripts/refresh-corpus.sh) to search the article database and handle updates. These commands are integral to the skill's stated purpose and are executed locally within the agent's environment.\n- [EXTERNAL_DOWNLOADS]: To keep the local corpus current, the skill fetches a JavaScript bundle from drli.beaucare.org. This download is limited to the specific domain and is used only as a data source.\n- [SAFE]: The skill exhibits several security best practices: it uses a custom-built parser to extract JSON data from remote scripts instead of executing them, verifies content using SHA-256 hashes, and uses a journal-based transactional system to prevent file corruption during updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 09:22 AM
Security Audit — agent-trust-hub — binlee-source-library