warp-deploy-nexus

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

• Third-party content exposure detected (high risk: 0.75). The skill may fetch Linear ticket URLs at runtime and then extract the warp route ID from the fetched ticket body/content, which is outsider-authored free text (Linear issue text) ingested into the agent context via the “fetch each ticket to extract the warp route ID” step.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs the agent at runtime to fetch provided Linear ticket URL(s) (the "Linear ticket URL(s)" referenced in the Input and Step 1) to extract warp route IDs, and those fetched values directly determine what the agent will add to the whitelist and include in the PR content.