tensorlake

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

• Third-party content exposure detected (high risk: 0.95). The required runtime workflow performs WebFetch https://docs.tensorlake.ai/llms.txt and then WebFetch <page>.md from the live docs index, which ingests outsider-authored free text from public web content into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires fetching the live docs at runtime (WebFetch https://docs.tensorlake.ai/llms.txt and then fetching the referenced .md pages from https://docs.tensorlake.ai/...), and instructs the agent to use those fetched pages as the authoritative source-of-truth that directly controls prompt content and generated code, so this is a runtime external dependency that influences agent instructions.