specialized-chief-of-staff
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's core functionality involves the agent ingesting and propagating information across various documents, which introduces a potential surface for indirect prompt injection attacks.
- Ingestion points: External documents, templates, sequences, assets, and project knowledge referenced in the 'Cascading Updates' and 'Output Routing' sections of SKILL.md.
- Boundary markers: The skill does not provide specific instructions to the agent to use delimiters or to disregard embedded instructions within the documents it processes.
- Capability inventory: The instructions involve implicit capabilities to read from and write to diverse project files to maintain a 'Document Dependency Map'.
- Sanitization: No sanitization, escaping, or validation steps are outlined for handling content retrieved from external sources before it is used to update other documents.
Audit Metadata