Offbrief — "production traffic is the only witness your agent didn't write"
Benchmark run · issue #23 · attempt 1 · MR !16 · phase BENCH_DONE
| seq | at (UTC) | step | record hash |
|---|---|---|---|
| 1 | 2026-10-15T12:09:45+00:00 | Issue #23 opened by @terrizoaguimor — the only human action | 96698d2d6497 |
| 2 | 2026-10-15T12:10:41+00:00 | The intent flow posted the contract note; the warden froze its sha256 1d5c7ec866d4 | 553b30840621 |
| 3 | 2026-10-15T12:10:47+00:00 | Merge request !16 linked to the builder session (bench) | 06df81921c0d |
| 4 | 2026-10-15T12:10:52+00:00 | The warden is about to create the sandbox pipeline (write-ahead record) | ff61f4fbda29 |
| 5 | 2026-10-15T12:10:56+00:00 | The warden started sandbox pipeline #2950778730 on head a60be86f, attempt 1 | aad84f1e2e5e |
| 6 | 2026-10-15T12:22:41+00:00 | Evaluation restarted (the previous pipeline was cancelled or failed) | 1c7310bfe574 |
| 7 | 2026-10-15T12:22:44+00:00 | The warden is about to create the sandbox pipeline (write-ahead record) | 482a3549b036 |
| 8 | 2026-10-15T12:22:47+00:00 | The warden started sandbox pipeline #2950822898 on head a60be86f, attempt 1 | e1522bbcd083 |
| 9 | 2026-10-15T12:31:11+00:00 | Evidence received from the pipeline (signed job token checked; pre-gate ok); evidence package sha256 41d024f025d7 | dbcc716260bf |
| 10 | 2026-10-15T12:31:15+00:00 | The warden posted the evidence note on the MR | 86ba07f93a22 |
| 11 | 2026-10-15T12:31:17+00:00 | The warden asked GitLab to start the judge flow (write-ahead record) | 43ad3486884e |
| 12 | 2026-10-15T12:31:19+00:00 | judge flow session started, run key offbrief-run:23:judge:1 | d46d760b8df8 |
| 13 | 2026-10-15T12:33:58+00:00 | judge flow session finished: finished | ab4fc8a317ec |
| 14 | 2026-10-15T12:33:59+00:00 | Judge verdict note: PASS | 72f44ef819ca |
| 15 | 2026-10-15T12:34:02+00:00 | Deterministic gate: all 12 checks passed | b75be9a0b40c |
| 16 | 2026-10-15T12:34:04+00:00 | Benchmark result: gate passed | c9ef819665d9 |
| check | result | reason |
|---|---|---|
budget_ok | pass | |
contract_ok | pass | |
controls_ok | pass | |
egress_ok | pass | |
evidence_ok | pass | |
head_ok | pass | |
judge_ok | pass | |
paths_ok | pass | |
pin_ok | pass | |
replay_ok | pass | |
scans_ok | pass | |
state_ok | pass |
1d5c7ec866d483a7375427714c6e3dcf04fd308e06cc79aa97bce3c572658b7341d024f025d766f22de92e36267e999fd7156a5c657fb56ef71108357582fdeb—sha256:5d9e51dbda428bcd94c26e0d566491af6954c7e851d099d24774b63753213ea7—c9ef819665d93dd11ad9dbd2fdcc378fd918284aa3a848c20e1db0da8a2abe43 (seq 16)From a clone of the repository, with the files above saved:
.venv/bin/python -m offbrief ledger verify --export 23.json
No install? The index has a ten-line, stdlib-only Python check.