Run your prompts
Choose agents and repeats. Each run starts a fresh conversation.
Available to the agents
Only Records MCPThese 4 tools are supplied in the illustrated configuration. An agent can choose a tool or respond without using one; run availability remains unknown where it was not recorded.
search_records · get_record · update_record · export_records
Agents
Each agent stores its model and harness. Use the same prompts across configurations.
| Agent | Configuration | Included |
|---|---|---|
| General assistant | Model A · Harness A | |
| Reasoning assistant | Model B · Harness A | |
| Fast assistant | Model C · Harness B |
This preview simulates progress with the illustrated Records fixture. It does not invoke models or your MCP.
Run cohorts are independent of production call volume. Tool invocation does not prove task success.