rag

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection.
  • Ingestion points: Untrusted content is ingested into the knowledge base through methods like add_document and add_file as seen in rag.py and SKILL.md.
  • Boundary markers: The context_template in rag.py provides minimal separation between context and user queries, lacking robust delimiters.
  • Capability inventory: Metadata in SKILL.md authorizes the use of system tools such as Bash, Read, Write, and Grep, which could be targeted by an injection attack.
  • Sanitization: The implementation in rag.py does not include any sanitization or filtering of the retrieved content before it is passed to the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 04:12 AM