slack

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

• Third-party content exposure detected (medium risk: 0.30). The skill integrates with Slack, which provides team messaging and feeds where outsider-authored messages can be read via API calls (such as conversations.history or search.messages) without the agent selecting a specific item beforehand.

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill includes an operational dependency on an external npm package (playwright-chromium) fetched and installed at runtime via npm install inside scripts/slack-browser-session.sh from the official npm registry, which is common/reputable unofficial publisher provenance.