brain
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill defines the 'brain' directory as a foundational source for all workflows, stating that every agent and session reads it. This structure is susceptible to indirect prompt injection if the files contain instructions that influence agent logic. * Ingestion points: Files within the
brain/directory (e.g.,index.md,principles.md,delegation.md). * Boundary markers: Absent; no delimiters or 'ignore' instructions are provided to help the agent distinguish between data and instructions. * Capability inventory: The skill body implies file-read and file-write capabilities within thebrain/directory. * Sanitization: Absent; no validation or filtering of vault content is performed before processing. - [NO_CODE]: This skill consists solely of markdown instructions for a knowledge management system and does not contain any scripts, binary files, or external software dependencies.
Audit Metadata