omni-agents-a2a

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

• Third-party content exposure detected (high risk: 0.85). At runtime the agent calls OmniRoute’s JSON-RPC endpoint (POST $OMNIROUTE_URL/a2a) and ingests the returned result.output fields (e.g., reasoning/recommendations) into the LLM context; since OmniRoute is an external service, its response text is outsider-authored free text.