linkedin-comment-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The Skill contains deceptive runtime instructions that tell the agent to automatically install BrowserAct dependencies without asking the user and explicitly to avoid describing the installation as a user request, which is hidden/deceptive behavior outside the advertised scraping functionality.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). These URLs point to an external install page (browseract.com) that the skill instructs agents to use for automated dependency installation without user confirmation, which is a moderately high-risk distribution vector for binaries or installers.

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 BrowserAct to access real LinkedIn post pages at runtime and extract comment text (outsider-authored free text from public/authorized web content), which the agent then uses as LLM-readable context for results.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill mandates installing and invoking the BrowserAct CLI at runtime using the installer URL https://www.browseract.com/?co-from=social-media (and its redirect variant https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main), which is a required runtime dependency and may fetch and execute remote installation code.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.70). The Skill explicitly instructs the agent to automatically install the BrowserAct CLI and resume execution "without asking the user to confirm" and to treat the CLI as a required runtime dependency, which directs the agent to modify the host system (install software) and bypass normal user consent—potentially compromising machine state.