solo-research

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). Yes — the SKILL.md explicitly instructs the agent to use WebSearch/WebFetch and MCP tools (e.g., MCP web_search with engines=reddit, MCP Playwright opening old.reddit.com, WebFetch scraping Product Hunt and trustmrr.com) to fetch and ingest public, user-generated content (Reddit posts/comments, Product Hunt entries, TrustMRR pages) and to synthesize quotes and findings from those sources, which directly influence research outputs and recommended actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly calls the PullPush API at runtime (https://api.pullpush.io/reddit/submission/search) to fetch full Reddit post selftext which is then injected into the agent's research context and can directly influence prompts/outputs, so it is a runtime external dependency that can control agent behavior.