Definitions
| Term | Definition |
|---|---|
| Target | The availability objective below, measured over a rolling 30-day window. |
| Measurement | Independent synthetic probe, evaluated continuously against the live endpoint. |
| Live status | status.mnemom.ai — the same signal these targets are measured against. |
| Breach | A sustained miss opens an incident on status.mnemom.ai automatically. |
Commitments
| Scenario | Definition | Target | Window |
|---|---|---|---|
| Platform availability | mnemom.ai and the core API respond | ≥ 99.9% | Rolling 30 days |
| Sign-up & login | Sign-up or login completes end to end | ≥ 99.9% | Rolling 30 days |
| Agent claim | A claim request records ownership | ≥ 99.9% | Rolling 30 days |
| AI gateway (LLM inference) | An authenticated inference request reaches the model | ≥ 99.9% | Rolling 30 days |
| Trust & reputation read | A reputation or trust-score read returns a current value | ≥ 99.9% | Rolling 30 days |
99.9% over 30 days ≈ 43 minutes of downtime budget per month, per scenario.
Review
| Cadence | Reviewed regularly. |
| Relaxation | Requires a published rationale on this page. |
| Live status | status.mnemom.ai |
