prospecting
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves mining data from untrusted public sources such as Reddit, Hacker News, GitHub issues, and various business websites to identify 'demand signals'. This behavior exposes the agent to indirect prompt injection if those external sources contain malicious instructions designed to hijack the agent's logic. \n
- Ingestion points: Discovery phases in
references/demand-signals.md(Reddit, HN, social posts) andreferences/local-prospecting.md(business websites). \n - Boundary markers: The skill instructions rely on LLM-based qualification against an ICP checklist but do not specify technical delimiters to segregate untrusted content. \n
- Capability inventory: The agent can generate CSV files, display data in chat, and execute the bundled
github-prospects.jsCLI tool. \n - Sanitization: While
references/demand-signals.mdmentions escaping values for HTML reports, there is no broader instruction for sanitizing text ingested from the web. \n- [COMMAND_EXECUTION]: The skill instructions inreferences/saas-prospecting.mddirect the agent to execute a bundled Node.js CLI tool (tools/clis/github-prospects.js) using a shell command. While the tool is local to the skill, the execution includes parameters (like repository names) that may be derived from user input or external discovery.
Audit Metadata