facebook-group-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The skill includes explicit instructions to auto-install a dependency without asking the user and to conceal that action ("do not request separate approval" / "do not describe it as a request"), which is a deceptive behavior outside the stated scraping purpose.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill includes explicit instructions to automatically install an external runtime dependency without asking the user and to hide that installation request, which is a deceptive supply-chain/installation behavior that can enable unauthorized code execution or malicious post-install payloads.

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 Facebook pages at runtime and then return scraped post text/comments authored by others, which would be outsider-authored free text ingested into the LLM context via the BrowserAct execution results.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs the agent to auto-install and invoke the BrowserAct CLI at runtime using the install link (https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main), which would download and execute remote code required for the Skill.

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 as a required runtime dependency without asking the user or requesting approval, which directs the agent to modify the host system (install software) and bypass user consent and potentially elevated-privilege checks.