khe-brain-isb
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill instructions or content. The agent is strictly constrained to a retrieval-and-citation workflow using a local, trusted knowledge base.
- [COMMAND_EXECUTION]: The skill provides a local validation script (
scripts/brain-check.mjs) used to verify the agent's adherence to citation rules. This script performs standard file I/O operations using built-in Node.js modules and does not perform network requests or execute arbitrary shell commands. - [INDIRECT_PROMPT_INJECTION]: The skill processes user queries by retrieving data from a static set of local markdown files provided by the author. The instructions in
SKILL.mdstrictly limit the agent to paraphrasing these notes and require clear source attribution, effectively mitigating risks associated with processing adversarial content. - [PROMPT_INJECTION]: The system instructions focus on enforcing accuracy and objective reporting based on the provided corpus, rather than attempting to bypass safety protocols or override platform constraints.
Audit Metadata