b2b-contact-mining
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs agents to search for and ingest data from external sources including Google, Telegram, and Discord to identify contacts. This creates a surface for indirect prompt injection where malicious data from these external sources could influence the agent's task execution or output.
- Ingestion points: Web search results and community messages from Telegram and Discord as specified in
AGENTS.mdandSKILL.md. - Boundary markers: The instructions do not provide specific delimiters or warnings to the agent to ignore instructions embedded within the scraped data.
- Capability inventory: The skill utilizes the
email_validator.pyscript, which has file system read access and network connectivity for SMTP verification. - Sanitization: There are no specified sanitization or validation steps for the retrieved content before it is processed by the agent.
- [DATA_EXFILTRATION]: The skill facilitates the collection and aggregation of contact information, including names and email addresses, into structured formats. While this targets public business data, the aggregation of Personally Identifiable Information (PII) requires appropriate security handling by the user.
Audit Metadata