chief-of-staff
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process information from external sources, such as 'transcripts,' 'subagent reports,' and 'research notes,' into a persistent wiki used for long-term decision-making. Malicious instructions embedded in these external sources could influence the agent's 'tribal knowledge' or future strategic decisions.
- Ingestion points: Processes external research notes, transcripts, and subagent reports as described in the 'Wiki' and 'Subagents' sections of SKILL.md.
- Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions embedded within the ingested data.
- Capability inventory: The skill has the authority to coordinate subagents, use background agents, and maintain a persistent file-based wiki across sessions.
- Sanitization: There is no mention of sanitizing, escaping, or validating content retrieved from external sources before it is committed to the wiki.
- [COMMAND_EXECUTION]: The skill is instructed to perform all work through subagents and background agents, which implies the execution of tasks and commands to fulfill user-defined goals.
Audit Metadata