meridian-investor-research
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external startup URLs provided by the user using the
web_fetchtool. - Ingestion points: The skill takes a
startup_urlas primary input and fetches its content in Step 1 of the workflow. - Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the data fetched from the external website, which could allow a malicious website to influence the agent's behavior during the summary or email generation phase.
- Capability inventory: The agent has the capability to fetch web data and conduct searches. It does not have access to sensitive local files, administrative commands, or persistence mechanisms.
- Sanitization: The skill employs specific extraction goals (e.g., 'What the company does in plain English', 'Traction numbers') which provide a functional layer of filtering, but there is no programmatic sanitization of the external content before it is processed by the model.
Audit Metadata