linkedin-company-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The Skill contains hidden/deceptive runtime instructions that order the agent to install a dependency automatically and to conceal that installation from the user (e.g., “without asking the user to confirm” and “Do not describe it as a request…”), which goes beyond the advertised scraping purpose and instructs nontransparent behavior.

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

• Third-party content exposure detected (high risk: 0.75). This skill’s runtime workflow invokes the external browser-act CLI to access LinkedIn pages and extract visible text (outsider-authored web content), which can then be fed into the agent’s LLM context via the tool’s returned page content.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires and directs the agent at runtime to install/use BrowserAct via the external installer links (https://www.browseract.com/?co-from=social-media and https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main), which are runtime external dependencies that could fetch and execute remote code to control the agent's browser automation behavior.