turso

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

• Third-party content exposure detected (high risk: 0.85). The skill’s required runtime workflow is to fetch live documentation text via WebFetch("https://docs.turso.tech/llms.txt"), which is public web content authored by Turso (an outsider) and will be ingested as readable prose into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs the agent to fetch live docs at runtime via WebFetch("https://docs.turso.tech/llms.txt"), meaning externally-fetched content would directly control prompts/instructions and is required for operation.