wiki-research-research

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from both external web sources and local queue files.
  • Ingestion points: The agent ingests data from RESEARCH_QUEUE.md and uses intelli_extract to pull content from external URLs found via web_search results.
  • Boundary markers: The instructions do not provide explicit boundary markers or directives for the agent to distinguish between its instructions and the data retrieved from external sources, making it possible for embedded malicious instructions in web content to influence agent behavior.
  • Capability inventory: The agent possesses capabilities to perform web searches, extract data from remote web pages, and write persistent files to the local vault (e.g., wiki/ directory and RESEARCH_QUEUE.md).
  • Sanitization: There is no evidence of content sanitization, filtering, or validation performed on the data retrieved from external URLs before it is processed through the research loop.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 11:52 AM
Security Audit — agent-trust-hub — wiki-research-research