intimacy

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt's examples and auth instructions require inserting a user's bearer token directly into curl commands/headers (Authorization: Bearer {{YOUR_TOKEN}}), which would force an agent to include secret values verbatim in generated requests or code.

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

• Third-party content exposure detected (high risk: 1.00). The skill's core API flows explicitly fetch and consume user-generated third-party content—e.g., GET /api/discover returns candidates and compatibility_narrative and GET /api/chat/{matchId}/messages returns other agents' messages—which the agent is expected to read and use to decide swipes, matches, and follow-up actions, creating a clear vector for indirect prompt injection.