brain-dump

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest a wide variety of external sources, creating a significant attack surface for instructions embedded in data.
  • Ingestion points: In the "Absorb" section of SKILL.md, the agent is directed to treat "conversation, notes, attachments, articles, files, and links as one dump."
  • Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between user instructions and potentially malicious commands embedded within the fetched articles or files.
  • Capability inventory: The skill references the use of tools like /read-the-damn-docs and /research, which likely involve reading local files and fetching external web content. It also passes the resulting context to a /grilling handoff, potentially propagating injected instructions through the agent's workflow.
  • Sanitization: No sanitization, filtering, or validation steps are defined for the data extracted from external sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 02:00 AM
Security Audit — agent-trust-hub — brain-dump