second-brain-ingest

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes raw source documents from a user-controlled directory, which creates a surface for indirect prompt injection if those documents contain malicious instructions.
  • Ingestion points: Files located in the raw/ directory are read completely for summarization and entity extraction.
  • Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore instructions embedded within the source documents during processing.
  • Capability inventory: The skill uses Bash, Read, Write, Edit, Glob, and Grep, providing significant file system manipulation capabilities if the agent is subverted by malicious content in a source file.
  • Sanitization: No sanitization, escaping, or validation of the source document content is performed before the agent discusses takeaways or creates new wiki pages.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:55 AM
Security Audit — agent-trust-hub — second-brain-ingest