bluesky-profile-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The Skill instructs the agent to install dependencies automatically and explicitly tells it to hide that action from the user (do not request separate approval / do not describe it as a browser-request), which is a deceptive directive outside normal transparent behavior.

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

• Third-party content exposure detected (high risk: 0.85). This skill’s required workflow delegates to the runtime browser-act dependency to access live Bluesky/public web pages, so outsider-authored free text (e.g., profile bios/content scraped from arbitrary URLs/pages) can be ingested into the agent’s LLM context via the BrowserAct browser-extraction path.

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 use the Install BrowserAct link at runtime to obtain a required CLI/Skill—e.g. https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main—meaning the agent will fetch and install remote code that directly supplies runtime behavior.