chief-of-staff
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from a wide range of external and untrusted sources, which could be exploited to deliver malicious instructions to the agent.
- Ingestion points:
SKILL.mdspecifies reading from pull requests, issue trackers, transcripts, sub-agent reports, development/production logs, databases, and browser content. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions to wrap external content.
- Capability inventory: Based on the instructions, the agent is expected to write to a local wiki (file system), query databases, read logs, and use a browser.
- Sanitization: No sanitization or validation processes are mentioned for the external data before it is recorded in the wiki or used for decision-making.
- [COMMAND_EXECUTION]: The skill orchestrates tasks by invoking other skills located in sibling directories (e.g.,
../retro/SKILL.md,../to-spec/SKILL.md). This behavior creates a dependency on the integrity of the surrounding file system and the security of those external instruction sets.
Audit Metadata