You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
tools.toml pmat pin 3.40.1 -> 3.41.1 — the fleet converged 2026-09-18 (infra#692) and every converged runner now fails check_tool_versions.sh #3490
pmat 3.41.1 is on crates.io (2026-09-18T14:48:52Z, after paiml/paiml-mcp-agent-toolkit release.yml run 35351362363's gate / cpu-gates went green on 516305ef0). infra#692 moved stack-tool-pmat to 3.41.1 on all five fleet boxes and converged them — including /opt/ci-tools/pmat, the file every container runner on gx10 and yoga mounts as /usr/local/bin/pmat, which nothing had declared and which was still 3.40.1 after the login user's pmat moved.
tools.toml says, in its own header, "Bump this file in the SAME commit as the infra forjar.yaml version bump." It is at 3.40.1. Measured in a fresh clone at 3409b29 with the converged pmat on PATH:
That is the verdict every aprender PR gets on every converged runner from now until this lands — the same split #3301 fixed for 3.40.1 (FAIL pmat pinned 3.40.0 found 3.40.1, a PR's verdict a function of which box picked it up).
What 3.41.1 carries that matters here
LockfileGuard (pmat#1406): pmat analyze dead-code no longer rewrites the analysed project's Cargo.lock under an ambient [patch.crates-io].
pmat roadmap aggregate (pmat#1364): the port of scripts/lib/roadmap_fragments.py aggregate; a follow-up PR replaces the roadmap-aggregate make targets with it once this pin is in, with byte-identical output proved (not this ticket).
Acceptance criteria
tools.toml[pmat] version = "3.41.1" and its header line naming the infra commit and PR; check_tool_versions.sh reads PASS on all rows with 3.41.1 on PATH.
CHANGING THE INSTRUMENT RE-RECORDS THE BASELINES (fix(tools): pin pmat 3.40.1 and CB-200 back to 602 — the fleet converged 2026-09-14; re-measuring under 3.40.1 read 603 (PMAT-3300) #3301): scripts/complexity_baseline.txt regenerated with check_complexity_ratchet.sh --update under 3.41.1; scripts/cb200_baseline.txt and .pmat-gates.toml [tdg] baseline re-measured with pmat comply check under 3.41.1 from a fresh index, and lowered if the reading is lower — never raised. If the reading is higher, the cause is named (instrument vs tree) before anything is written.
pmat_bin.sh's PMAT_PIN="3.37.0" and the fleet's pmat-337-pin stay: that is a different contract with a different reason.
The ticket's roadmap row is a fragment under docs/roadmaps/entries/, and make roadmap-aggregate-check is green.
Refs #3300 (the 3.40.1 bump, #3301), infra#690, infra#692, pmat#1408.
Taking this now (session 9e561dfb): tools.toml → 3.41.1 and the two baselines re-recorded under 3.41.1 in the same commit (check_complexity_ratchet.sh --update, cb200_baseline.txt), per the file's own CONVERGE → PROVE → ASSERT rule — the boxes are proven (intel's tool-probe line reads pmat=3.41.1 (/opt/fleet-bin/bin/pmat) on run 35363356002; lambda measured 3.41.1). It blocks #3489 (ONT-4) and every other open PR. If another session already has a branch for this, say so here and I will close mine.
Activity
noahgift commentedon Sep 18, 2026
Taking this now (session 9e561dfb):
tools.toml→ 3.41.1 and the two baselines re-recorded under 3.41.1 in the same commit (check_complexity_ratchet.sh --update,cb200_baseline.txt), per the file's own CONVERGE → PROVE → ASSERT rule — the boxes are proven (intel'stool-probeline readspmat=3.41.1 (/opt/fleet-bin/bin/pmat)on run 35363356002; lambda measured3.41.1). It blocks #3489 (ONT-4) and every other open PR. If another session already has a branch for this, say so here and I will close mine.chore(ratchet): CB-200 601→599 and the complexity shrink under pmat 3…