marketing-prospecting
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill processes external data from untrusted sources such as forums, social media, and third-party websites to extract pain signals and prospect info. * Ingestion points:
references/demand-signals.mdandreferences/local-prospecting.mdinstruct the agent to read content from public threads and business sites. * Boundary markers: The skill employs structured phases and qualification checklists to handle data. * Capability inventory: The agent has access to network-based research tools and local file creation (CSV). * Sanitization: Instructions explicitly mandate escaping all values when generating HTML reports. - [EXTERNAL_DOWNLOADS]: Integrates with well-known B2B data providers and development tools. The skill utilizes established platforms including Apollo, ZoomInfo, Clay, GitHub, Firecrawl, and Browserbase for data enrichment and site research. These references are to official services and do not involve untrusted remote code execution.
- [COMMAND_EXECUTION]: The skill includes instructions to run a local CLI tool for GitHub data processing. The
github-prospects.jsscript is used with static arguments to enrich prospect lists from repository stargazers, which is consistent with its stated purpose.
Audit Metadata