openclaw

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

• Third-party content exposure detected (high risk: 0.90). The skill's Research Process explicitly instructs a sub-agent to fetch the public documentation index at https://docs.openclaw.ai/llms.txt and to "fetch multiple relevant pages in parallel using WebFetch," so the agent ingests and acts on open/public third-party docs that could alter its actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs the sub-agent at runtime to fetch the documentation index (https://docs.openclaw.ai/llms.txt) and then retrieve docs that will be injected into the agent's research context and used to form prompts/instructions, so https://docs.openclaw.ai/llms.txt is a required runtime external dependency that directly controls agent behavior.