zoho

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

• Third-party content exposure detected (medium risk: 0.65). At runtime, the skill uses ClawLink to perform Zoho “actions” by browser/OAuth and then fetches Zoho action metadata (e.g., via actions zoho / describe zoho) and possibly Zoho record content; that content is outsider-authored (Zoho tenant data and Zoho-provided text) and is ingested into the agent’s LLM context through the CLI’s readable outputs.

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 run "npx @useclawlink/cli" which fetches and executes remote JavaScript from the npm package (https://www.npmjs.com/package/@useclawlink/cli / https://registry.npmjs.org), so a runtime external dependency will execute remote code that controls the agent's actions.