skills/openai/plugins/pharmgkb-skill/Gen Agent Trust Hub

pharmgkb-skill

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEPERSISTENCEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [Potential for Arbitrary File Write]: The script scripts/rest_request.py provides a raw_output_path parameter that allows the agent to save the content of a network response to any specified file path. This capability could be used to write data to sensitive system files or configuration locations if the tool is misused.
  • [Unrestricted Network Access]: The tool is designed to query PharmGKB but uses a flexible base_url parameter that is not restricted to official domains. This allows the script to be used to make network requests to any reachable server, which is a consideration for Server-Side Request Forgery (SSRF).
  • [Indirect Prompt Injection Surface]: The skill processes data from external APIs and possesses capabilities to write files and perform network requests. Ingestion points: API responses processed by scripts/rest_request.py. Boundary markers: The instructions do not define delimiters for external data. Capability inventory: The script scripts/rest_request.py can perform network requests and file system writes. Sanitization: There is no validation of the target file path or sanitization of the response content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 12:20 PM
Security Audit — agent-trust-hub — pharmgkb-skill