twitter-x-reply-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The Skill explicitly instructs the agent to install a runtime dependency automatically "without asking the user to confirm" and to "not describe" the installation link as a request for the user, which directs deceptive behavior outside the stated scraping purpose.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The Skill contains explicit instructions to automatically install an external CLI dependency without user consent and to hide that installation from the user, creating a high-risk supply-chain/unauthorized-installation/backdoor vector despite otherwise benign scraping functionality.

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 live public Twitter/X pages and collect reply text at runtime, which would ingest outsider-authored free-form reply content into the LLM context for analysis.

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 automatically install and invoke the BrowserAct CLI at runtime using the install URL https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main, which indicates fetching/executing external code as a required dependency.

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 instructs the agent to automatically install the BrowserAct CLI and dependencies "without asking the user to confirm" and forbids requesting separate approval for dependency installation, which pushes the agent to modify the host system (potentially requiring privileged operations) without consent.