recoup-internal-consulting-comms-logger
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from emails, LinkedIn messages, and screenshots which could contain malicious instructions intended to manipulate the agent.
- Ingestion points: In Step 2, the skill captures "As-sent / as-received text, verbatim" and engagement signals from external communication channels.
- Boundary markers: There are no explicit delimiters or instructions to ignore embedded prompts within the captured message text.
- Capability inventory: The skill has the ability to write to repository files (AGENTS.md, _board.md, followups.md), query Gmail threads, and modify data in the Attio CRM.
- Sanitization: The instructions do not specify any sanitization, filtering, or escaping of the external message content before it is recorded or used to update deal states.
Audit Metadata