solo-community-outreach
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external platforms including Reddit, Hacker News, and ProductHunt via
WebSearch,WebFetch, and custom MCP search tools. This external content is interpolated into prompts to generate outreach drafts and a summary plan written todocs/outreach-plan.md. There is no evidence of strict boundary markers or sanitization for this external data, creating a surface where malicious instructions in community threads could influence the agent's output. - [EXTERNAL_DOWNLOADS]: The documentation references a GitHub repository
github.com/fortunto2/searxng-docker-tavily-adapteras a recommended self-hosted search adapter. This repository is a vendor-owned resource matching the skill author 'fortunto2'.
Audit Metadata