Releases: bsebench-org/bsebench-async-codex
v1.0.0-rc2 — sister-repo aligned
v1.0.0-rc2 — 2026-05-20
Sister-repo release candidate aligned with bsebench-runner v1.0.0-rc2.
This tag fast-forwards main to demo-k's HEAD (d100ff2) for the BSEBench v1.0 release candidate. See the full change-log, the operator-track residue, and the source pins for all sister repos at:
https://github.com/bsebench-org/bsebench-runner/releases/tag/v1.0.0-rc2
This is a release CANDIDATE, not a final v1.0 release. The release event (Zenodo DOI, GPG/Sigstore signature, ≥4 Steering-Committee seats, S3 trifurcation) is still on the operator track.
M1 — Governance unchanged from M0
M1 made no changes to governance. The bsebench-async-codex demo-k branch and GOVERNANCE.md v0.1 carry over from M0.\n\nFull release: https://github.com/bsebench-org/bsebench-runner/releases/tag/m1
M0 — Governance v0.1
bsebench-async-codex @ M0
Governance artifacts for BSEBench M0 (Foundation Hardening).
Full release context: https://github.com/bsebench-org/bsebench-runner/releases/tag/m0
Branch policy
This repository's main branch belongs to Codex (the candidate's separate AI implementer, parallel timeline) and remains untouched by M0.
M0 ships its governance artifacts on a new demo-k branch off main. The tag m0 is on demo-k, not on main. Codex's Phase 56 baseline on main is preserved as the parallel timeline.
Shipped at this tag
GOVERNANCE.md v0.1 — 10 sections
- Mission — protocol-layer SSoT, survival of candidate graduation.
- Steering Committee structure — 7 seats: 3 institutional (battery aging, characterization, degradation modeling) + 2 industry + 1 methods + 1 candidate (ex officio until defense). Vacant-seat policy (operational with ≥ 4 filled). Staggered terms.
- Decision-making — 4 thresholds: routine (1 reviewer), minor (3 of 6), breaking (4 of 6), composition (4 of 6). Candidate veto until defense.
- RFC process — 14-day discussion (minor) / 28-day (breaking) + 7-day vote. Sigstore-signed once M5 ships.
- Sustainability plan — GitHub Sponsors immediate; 3 grant tracks (ARPA-E SCALEUP, EU Battery 2030+, DOE-VTO); industry consortium at month 12 post-v1.0 iff ≥ 3 industry submissions (no voting rights for sponsors).
- Conflict of interest — recusal rules, annual disclosure.
- Adoption KPIs — 4 KPIs with at-risk thresholds, calibrated to the "30 datasets mostly grade C" honest framing.
- Code of conduct — Contributor Covenant 2.1. Phased reporting channel (per advisor §4 R9 fix): Mme Rim institutional email M0.5 → M4, then private GitHub repo +
conduct@bsebench.orgat M5. - Candidate exit protocol — 30-day succession RFC, credential downgrade, candidate seat becomes community-elected.
- Amendment process — governance is itself versioned and audited.
claim_status of governance v0.1: NO_CLAIM until ratified by the first Committee meeting.
3 invitation drafts — NOT SENT
steering_committee/invitations/draft/_NOT_SENT_README.md— send checklist (Mme Rim co-sign, PDF rendering, institutional email, 21-day window).steering_committee/invitations/draft/nasa_pcoe_invitation_draft.md(Battery Aging seat).steering_committee/invitations/draft/calce_invitation_draft.md(Battery Characterization seat).steering_committee/invitations/draft/oxford_invitation_draft.md(Degradation Modeling seat, PyBaMM F3 dependency acknowledged).
Each letter: ~2h/month, RFC voting, quarterly call, attribution & citation discipline in return.
Sending these letters is gated on Mme Rim's co-signature + Oussama's explicit go-ahead. The drafts are versioned audit-trail evidence that M0 governance ships real artifacts, not placeholders.
Read alongside
- Main release notes: https://github.com/bsebench-org/bsebench-runner/releases/tag/m0
- Authoritative roadmap §F7 (Governance from Day Zero):
bsebench-runner/docs/SSOT_BSEBENCH_ROADMAP_AUTHORITATIVE_2026-05-12.md - Advisor challenge §4 R9 (conduct channel) + §2 Q5 (governance realism critique):
bsebench-runner/docs/ADVISOR_CHALLENGE_M0_2026-05-12.md