skills/openai/plugins/rcsb-pdb-skill/Gen Agent Trust Hub

rcsb-pdb-skill

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [Generic REST Client Usage]: The scripts/rest_request.py script acts as a general-purpose REST client. While the skill's instructions focus on the RCSB PDB service, the underlying script is capable of performing HTTP GET and POST requests to user-defined endpoints with custom headers and bodies.
  • [File System Interaction]: The skill includes functionality to save raw API responses to the local file system through the save_raw and raw_output_path parameters. This allows for data persistence and is intended for scientific workflows where raw JSON or FASTA files are required.
  • [Data Retrieval from External Services]: The skill fetches configuration and data from rcsb.org, a well-known scientific database for structural biology. These network operations are essential for the skill's primary function of retrieving protein structure metadata.
  • [Indirect Prompt Injection Surface]: The skill processes data from external API responses, which introduces a potential surface for indirect instructions.
  • Ingestion points: External data enters the agent context through the execute function in scripts/rest_request.py via network calls.
  • Boundary markers: The script returns data in a compact format, but it does not utilize explicit markers to isolate retrieved content from the agent's instructions.
  • Capability inventory: The skill includes network access and file-writing capabilities across its scripts.
  • Sanitization: The script applies data truncation and compaction to manage response size, though it does not perform deep content sanitization for embedded instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 12:21 PM
Security Audit — agent-trust-hub — rcsb-pdb-skill