erc8004-agent-layer

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

• Third-party content exposure detected (low risk: 0.10). The skill’s runtime path ingests only user-supplied partner manifest JSON (e.g., provider/assessmentSource/entity fields) and executes read-only GraphQL preflight via scripts/build-partner-plan.mjs → resolveAgentPreflight() and preflightQuery(); it does not follow attacker text beyond those structured fields and does not fetch arbitrary outsider-authored web content.