resolve-before-asking

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting data from untrusted external sources, which creates a potential surface for indirect prompt injection attacks.
  • Ingestion points: Data enters the agent's context through gbrain query (which accesses email archives, chat transcripts, and calendar imports) and external web search results.
  • Boundary markers: The instructions do not mandate the use of delimiters or specific safety instructions to isolate external content from the agent's core logic.
  • Capability inventory: The skill utilizes the gbrain CLI suite for data synthesis and possesses write permissions for the people/ and companies/ directories.
  • Sanitization: There is no mention of sanitizing or filtering external content before it is synthesized or written to the internal database.
  • [COMMAND_EXECUTION]: The skill provides instructions for using a specific suite of CLI tools to perform complex data lookups and graph operations.
  • Evidence: Documented use of gbrain think, gbrain search, gbrain get, gbrain query, gbrain sources, gbrain timeline, gbrain backlinks, and gbrain graph to synthesize information from the internal database.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 06:18 PM
Security Audit — agent-trust-hub — resolve-before-asking