prospecting
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill references the use of a local command-line script,
tools/clis/github-prospects.js, to extract and enrich lead data from GitHub stargazers and forks. - [EXTERNAL_DOWNLOADS]: The workflow relies on multiple external B2B data providers and APIs, including Apollo, ZoomInfo, Clay, Clearbit, and Truelist, for firmographic enrichment and email deliverability verification. It also leverages Firecrawl and Browserbase to fetch content from individual business websites.
- [PROMPT_INJECTION]: The 'Demand-signal' prospecting branch involves ingesting and analyzing untrusted content from public sources like Reddit, Hacker News, and various online forums.
- Ingestion points: Data is retrieved from community threads, social media posts, and public reviews via web search and scraping tools.
- Boundary markers: The agent is instructed to paraphrase and quote minimally, but the instructions do not implement specific delimiters to separate untrusted data from the system prompt.
- Capability inventory: The skill possesses the ability to write collected data into local CSV files and perform network requests to several enrichment APIs.
- Sanitization: No specific filtering or validation logic is defined for content mined from external discourse before it is processed by the agent for scoring and outreach drafting.
Audit Metadata