linkedin-founder-lead-search

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The prompt contains deceptive runtime instructions that require the agent to install dependencies automatically without user confirmation and to conceal that action (e.g., "do not describe it as a request" / "do not request separate approval"), which overrides transparency and falls outside the advertised lead-search purpose.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md instructs the agent to use the BrowserAct CLI to access and extract live public LinkedIn pages at runtime, which would ingest outsider-authored free text (e.g., profile/post content) into the LLM context via the browser-extracted page text.

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

• Potentially malicious external URL detected (high risk: 0.90). The Skill explicitly instructs the agent to automatically install the BrowserAct CLI at runtime using the install link https://www.browseract.com/?co-from=lead-generation&redirect=https://github.com/browser-act/skills/tree/main, which implies fetching and executing remote installer code as a required dependency.