knowledge-acquisition

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates discovery and retrieval of data from academic APIs and developer platforms. It fetches paper metadata and PDFs from arXiv, Semantic Scholar, OpenAlex, and CrossRef, and clones source code from GitHub repositories.
  • [COMMAND_EXECUTION]: The skill uses local CLI tools for operational tasks. It calls git for repository management, gh (GitHub CLI) for metadata searches, and npm/node for building and running the web visualization application.
  • [SAFE_PRACTICE]: Includes a dedicated url_utils.py module that enforces protocol safety by validating URL schemes (restricting them to http/https) before opening network connections, effectively preventing potential SSRF or protocol-handler attacks.
  • [SAFE_PRACTICE]: The included AWS infrastructure code follows security best practices by utilizing Cognito for user authentication, IAM roles for agent permissions (least privilege), and S3 Origin Access Control (OAC) for the CloudFront distribution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 03:49 AM
Security Audit — agent-trust-hub — knowledge-acquisition