agent-network-a2a

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly discovers and calls arbitrary agents on the open agents.hot A2A network (see SKILL.md Step 1 "agent-network discover" and Step 3/4 "agent-network call" and the chaining example that saves /tmp/trend.txt and feeds it to another agent), thereby ingesting untrusted, user-published agent responses which are read and used to drive subsequent calls and decisions.