task-writer

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

• Third-party content exposure detected (high risk: 0.80). The skill's required Task Writing Workflow explicitly instructs the agent to check and read the public Task Library (https://tasks.mechanic.dev) and other external docs (also lists GitHub, learn.mechanic.dev) and the included patterns show using HTTP actions to external endpoints, meaning the agent is instructed to fetch and interpret untrusted public third‑party content that can influence its subsequent actions.