# Evolution 80 — Durable Execution Report

> Generated by `scripts/evolution80/docs_e80.py` from `artifacts/e80/` (run 3). FIRME AI CORPORATION — a three-person team (founder: Vincent Noriega).
> Development branch `evolution-80`. Not deployed. No cloud model was called. No third-party audit.

Scope: single host, 4 worker processes. This is a one-host failure-injection drill, not a distributed deployment.

| injected failure | injected |
|---|---|
| worker_sigkill_mid_task | True |
| worker_sigkill_mid_task | True |
| worker_sigkill_mid_task | True |
| stale_worker_frozen_past_lease | True |
| coordinator_sigkill_restart | True |
| storage_write_lock_5s | True |
| interrupted_tool_execution | True |
| interrupted_model_inference | True |

- Tasks 44; done 44; retried 7; max attempts 2.
- External effects 43 with 43 distinct idempotency keys; **duplicate effects 0**; duplicates suppressed 1.
- Stale commits refused by fencing: 2. Coordinator starts: 2.
- Completions recomputed 41; mismatches 0; done without an effect record 0.
- Wall time 65.2 s.

## Not evaluated in Evolution 80

| capability | status |
|---|---|
| checkpoint_replication | NOT_EVALUATED in E80 |
| cross_host_scheduling | NOT_EVALUATED in E80 |
| host_reboot | NOT_EVALUATED in E80 |
| network_partition | NOT_EVALUATED in E80 (single host); E72/E78 evidence preserved, not re-run |
| regional_scheduling | NOT_EVALUATED in E80 |

Cross-host scheduling, checkpoint replication and network-partition handling were measured in Evolutions 72 and 78; that
evidence is preserved and was not re-run here.
