Repository navigation
contractiles: state the Mathesis contracts (re-anchoring WO 3.3) - #76
Conversation
The owner reviewed and approved every clause through the selector on 2026-11-11 (WO 3.3 needs a ruling: the contractiles are a sensitive area). - intend: the purpose, anti-purpose and doubt texts describe Mathesis. The template intents and wishes give way to the six stages of the re-anchoring plan: stages 0 to 2 done, each with a probe; stage 3 in progress; stages 4 and 5 declared. - must: six critical invariants in a section "Mathesis Invariants": views-current, no-private-name, no-hand-typed-grade, receipts-stay-home, home-theorem-has-receipt, rows-have-provenance. Each probe is offline and prints no private name. The abstract says that no runner executes must probes yet; the CI gate in each note enforces the rule. No template clause is removed: each can apply. - trust: a section "Writers" says who writes which record. - witness map: the entry for the removed intent repo-initialization goes. No deed fact attests a stage, so no new witness is added (owner ruling 2026-11-11: probe-deferred is the expected reading). - coaptation receipt regenerated. It reads amber: three committed stages are open and the tolerance is two. The owner accepted that reading. Observed in this session on main bd5259a plus this change: the six must probes and the three intend probes pass; each must probe fails on a planted fault (stale catalogue, private frame record, PWG in a pin, a second receipt deed, the THY-1 pin removed, a row without its commit); the k9 validator (standards 8e48b1c), the canon deed lint, just validate-coapt, validate-claude-md, validate-state, validate-rsr and validate-repo-map pass. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0136P7Zo9dheFQzYYWV49hAk
|
Caution Review failedThe pull request is closed. Walkthrough
Merge Risk:
 |
|
Codex Review Summary: ❌ FailedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Summary
Work order 3.3 of
docs/proposals/mathesis-re-anchoring.adoc. The contractiles are a sensitive area, so WO 3.3 needed a ruling. On 2026-11-11 the owner ruled through the selector: follow the plan, show the text first, run WO 3.5 before WO 3.3, use probes and accept "probe-deferred", severitycritical, accept an amber band. The owner reviewed and approved every clause below before this commit.Changes
intend: the purpose, anti-purpose and doubt texts describe Mathesis. The template intents and wishes (repo-initialization and six more) give way to the six stages of the plan. Stages 0 to 2 are done, each with a probe. Stage 3 is in progress. Stages 4 and 5 are declared.must: six critical invariants, in a section "Mathesis Invariants":views-current,no-private-name,no-hand-typed-grade,receipts-stay-home,home-theorem-has-receipt,rows-have-provenance. Each probe works offline and prints no private name. The abstract now says that no runner executesmustprobes yet; the CI gate named in each note enforces the same rule. No template clause is removed: each one can apply to a minted repo.trust: a new section "Writers" says who writes which record.repo-initializationgoes. No deed fact attests a stage, so no witness is added.RSR Quality Checklist
As Applicable
.a2mlfile, no private nameTesting
Run in this session on
mainbd5259aplus this change:mustprobes and the threeintendprobes: passmustprobe, each caught: stale catalogue, private frame record,PWGin a claim pin, a second receipt deed, the THY-1 pin removed, a row without its commitstandards8e48b1c, the CI pin): 14 files, 0 failingjust validate-coapt,validate-claude-md,validate-state,validate-rsr,validate-repo-map: passNot run here:
just check-proofs(no Guix and no provers in this container; CI runs it).Finding: the generated coaptation receipt holds one U+2014 character. It comes from the advice text in
.machine_readable/coaptation/coapt.ncl, not from this change, and a hand edit would breakvalidate-coapt.Generated by Claude Code
Summary by CodeRabbit