docs/analysis/floor-memory-profile-2026-10-08.md: the 3060 and 4060 rows on the default, unoverridden job path (the served 0317 host on the V6-07 server: 3060 shard 11.4 s at 7,601 MiB, the whole chain 27.4 s at an 8,306 MiB aggregation peak, both VERIFIED; 4060 shard 16.3 s at 7,504 MiB VERIFIED, the chain aborts at the aggregation's 486 MiB allocation on 176 MiB free), the host's profile and refusal rows (exit 78 under the floor and beside the miner on both cards, the miner unharmed), the recursion constant isolated (no change of the device peak), the known-failed first server, the master-host 0-bytes fault with its NOT RUN rows, and the consequences per card tier. memory_profile.rs: aggregate and chain floors 8,500 MiB (an 8 GB card proves shards and is refused the chain before setup), tests re-pinned; suite 7 passed on box 3.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The window lane's engine half (reviewb-202 af5ade43) grants a fixed proof budget beside the miner (7,532 MiB plus 10 percent) while the card's free figure can be under it (a 12 GB card beside the 6.1 GiB miner: 6,159 MiB free), so memory_profile::lease_reading takes the lesser of IGNEUM_PROVE_MEM_BUDGET_MB and IGNEUM_PROVE_MEM_FREE_MB when both are set; test lease_takes_the_lesser_of_grant_and_free. Suite on box 3: 7 passed.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Documents only: docs/analysis/class-v6/amd-intel-energy.md and docs/analysis/class-v6/amd-intel-energy/ (the three job scripts, the Level Zero sampler source, the offline-compile prelude and ISA summary).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Documents-only replay of d9e91c5d7 (d9e91c5d7a) for the box mirror master
tools/fleet/box-prover.py point 6: each claimed segment is an eligible job and closes once as paid, expired (unpaid, held_expired) or cancelled (disk, export, cut, chain, timeout, shards, statement, sign, refused), active until then; the row carries outcome, cause, deadline, margin at the claim, seconds spent (wasted unless paid) and the deadline miss; the state carries the counters and the wasted seconds by cause; every close is a RESULT outcome line, the run ends with a RESULT ledger line.
tools/fleet/prover-outcomes.py: the report over state files and logs (a log from before the ledger is reconstructed from its RESULT lines; a state row wins over a log row): outcomes, paid completions (segments, shards, IGN, median end to end, median time to pay, median margin), missed deadlines by cause with the median miss (never negative, none on a stale tip), wasted work by cause, throughput per hour and the paid share of the seconds spent, the open jobs. --self-test to known numbers (known-failed first), run by the gate. tools/fleet/night.py sums the fleet's counters into its hourly row. docs/analysis/proving-outcome-ledger.md: the definitions and how to read the three ratios.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Documents only. The master (292 pages) governs where it differs from the
plan PDF or the RTF; 18 INT gates and 44 R2 closure requirements become
suites INT and REV in the acceptance registry (CI steward).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
GPU-01, GPU-03 and POW-08 take run_status RUNNING with the evidence rows on build-1 under /srv/artefacts/tas/p02-rows; PASS waits on the AMD cells, the 1080 Ti cell and the re-exported v6 packs.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Documents only. Fourteen findings F01 to F14 with forty-four regressions,
dispatched by lane in dispatch.md; the regressions become suite REV in the
acceptance registry (CI steward).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The map gains harness:reference-apps-compat (EVM-02 to EVM-06, partial coverage named per case), harness:reference-apps-ver
(VER-03 to VER-06 and VER-08) and harness:reference-apps-ver-trust-anchor (VER-01 and VER-02, FAIL by design and disclosed).
Batches ra-20261008T1744-evm (22 of 22 rows on the 2.0 devnet, node 4cdcc488) and ra-20261008T1746-ver (evidence per case
under tools/reference-apps/ver/evidence) are recorded with node tools/ci/test-record.mjs --record; the harness map page is
regenerated; the registry rows move with their evidence (site/lc/verify-receipt.js rebuilt in this landing).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Run 201-ef0f2ed8-2826f37e: node ef0f2ed8 (291ee6ae plus the 2.0.1 version bump) on the key-succession pairing, kaspad check and the five node-side suites green on build-4; the miner cut tip 2826f37e (app crate identical to 2ea7b43f), pow and app green on build-2; evidence at build-1:/srv/artefacts/tas/201-ef0f2ed8-2826f37e/. GOV-01's row takes F0 (docs/plans/igneum-2.0-f0-manifest.md) as its evidence through the batch f0-manifest-20261008-b (the earlier f0 batch landed before the replay existed and never recorded). The map splits the finality cells at the finality lane's ask: harness:finality-sim keeps FIN-08 (the model side), harness:finality-sim-fin02 carries FIN-02's simulator half, node:finality-realnode carries the node lane's real-node lines for FIN-02 and FIN-07.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
