event-prospecting-pipeline
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill is designed to collect and export personal and professional lead data (names, titles, emails, LinkedIn profiles) to external services such as Google Sheets and the AgentMail API (agentmail.dev) for outreach purposes. This behavior is consistent with the primary prospecting function described in the playbook.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes untrusted data from the web.
- Ingestion points: Processes data from Luma event URLs, attendee bios, and conference speaker websites (SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between data and potential instructions within the ingested bios or site content.
- Capability inventory: Includes web searching, writing to spreadsheets/CSV, and initiating network requests via the AgentMail API (SKILL.md).
- Sanitization: No evidence of sanitization or validation of the external content before it is processed by the agent or qualification steps.
Audit Metadata