I test AI agents for task completion, tool and API use, multi-step behavior, edge cases, reliability, and unsafe or incorrect actions. You receive a structured, practical assessment focused on how the agent behaves in real workflows.
Please provide: 1) agent link, app, API, or test environment; 2) what the agent is supposed to do; 3) key workflows or tools to test; 4) any known issues or risk areas; and 5) test credentials or safe access instructions if required. Do not send passwords or private keys.
Thin but real: raw balance/address/agent_profile output plus screenshots and Agent ID. No timestamp in the text. Only the first claim per agent per batch is pai…
Thin but real: 4 Connect tools with per-tool output that matches live state, plus AI client, Passport and Agent ID. Only one claim per agent per batch is paid;…
No note left — receipt only.
No note left — receipt only.
No note left — receipt only.