bots.st/saagpatel
operant-mcp
Benchmark case and results retrieval
2172 ms
answered our call
- State
- Answered
- Protocol
- MCP 2025-06-18
- Tools listed
- 5
- Answered
- 3 of the last 3 calls
How to reach it
- MCP, HTTP
https://operant-mcp.saagar210.workers.dev/mcp - npm
npx -y saagar-operant-mcpReleased 18 Jul 2026, version 0.1.1
What it can do
| Tool | What it says it does |
|---|---|
compare_models | Place two retained calculation profiles side by side by display_name substring. Returns ocs_mean, ocs_stdev, orchestration_mean, run_family, and subject_shell for each, plus comparison_status=NOT_DURABLE and the public claim_status. This is not evidence that either named model outperforms, equals… |
get_case | Return the full case for a given pair_id (axes 1/2/4) or id (axis 3): malign and benign task prompts, expected decisions, grounding rationale, and bypass patterns. Axis 3 cases are single (unmatched) and use an 'id' field instead of 'pair_id'. Use list_cases to browse available ids. |
get_methodology | Return the benchmark design: the 4 axes, the OCS formula (Youden's J), the 5 decision labels (PROCEED, PROCEED_SANCTIONED, REFUSE, ESCALATE, REROUTE), scoring blocks, and a concise what-it-measures / what-it-doesn't summary. Sourced from the baked case-file metadata. |
get_results | Return retained model calculation profiles plus generated_at, claim_status, claims_at_risk, and the bounded evidence_binding. These rows are not durable named-model performance claims. Do not rank them or treat stdev/significance as reliable model evidence. |
list_cases | Return case metadata (no full task prompts): pair_id/id, axis, tier, grounding, and side indicators (malign/benign for axes 1/2/4; null for axis 3). Filter by axis, or omit for all cases across all axes (the result includes a count). Use get_case to fetch a full case with task prompts and expected… |
In its maker's words
Read-only MCP server for the OPERANT AI operating-agent calibration benchmark.
- Source
- github.com/saagpatel/operant
- Publisher
- saagpatel (6 listings in the registry)
- Listed from
- the official MCP Registry as
io.github.saagpatel/operant-mcp - Version
- 0.1.0, updated 28 Jun 2026
- Kinds of work
- AI models and agent tools, Search and research
Every call we made
| When | What happened | Time |
|---|---|---|
| Answered | 2172 ms | |
| Answered | 1990 ms | |
| Answered | 509 ms |
Next door
- WhiteMagic — governed local memory for agent continuityGoverned local agent memory reads and writes
- PolyrankTrader metrics, rankings, and strategy backtests
- DiffCIValidate DiffCI reports and plan affected tests
- FOGA nodeResolve agent identities and safety verdicts, simulate and r
- Mind Upload AI meta harnessRoute coding work through saved memories
Is this your bot? Claim its address to add a verified badge and keep the listing current.