Skip to content

Sat May 16 - #3287

Closed
briansrls wants to merge 10 commits into
mainfrom
session/still-hawk-102-scaffold-exception-reconcile
Closed

briansrls wants to merge 10 commits into
mainfrom
session/still-hawk-102-scaffold-exception-reconcile

Conversation

@briansrls

Copy link
Copy Markdown
Contributor

Auto-opened by session-dashboard for session still-hawk-102.
Pushing to session/still-hawk-102-scaffold-exception-reconcile advances this PR.

Worker attestation

Before flipping this PR to ready for review, confirm each item:

  • Title describes the change (not the session id or branch).
  • PR body summarises what and why (replace the TODO below).
  • Tests run: name the command (e.g. npm test, cargo test) and the result.
  • If this closes a work item, the body contains a Closes #N directive.
  • No commits on this branch are surprises (no fork/cherry-pick I did not make).
  • No secrets / credentials / large binaries staged.

Summary

TODO: replace this paragraph with one or two sentences naming the change and its motivation. Reviewers read this first.

Test plan

  • TODO: list the commands that ran (or "no tests changed; relied on CI") and the outcome.

briansrls and others added 10 commits May 18, 2026 03:27
…🟡 plan-binding

BLOCKING codex review (#3244, sha 15242ef + inline :257): #3244 made 🟡
plan-binding universal (every 🟡 binds a named primitive + owning
substrate PR + dissolve-on-arrival follow-up), but Practice 4's scaffold
exception still let the DECISIONS.md ledger defer wholesale and treated
`feature:<sunset-milestone>` itself as the gate. A milestone-only gate
leaves the dissolution path non-checkable, and a wholly-deferred ledger
has nowhere to record the plan-binding.

Fix: the scaffold exception now defers only the verbose
patterns-tried analysis — NOT the 🟡's bound dissolution plan, which is
the minimum that makes a 🟡 valid and is recorded even for a scaffold.
The gate names the concrete missing primitive; a bare sunset-milestone
is not a valid gate (the milestone records when revisited, not what is
awaited).

Audit also surfaced a minor gap: the Calibration section's general
NON-BLOCKING tier did not cross-reference that dissolution findings are
an always-BLOCKING class — added that line so a dissolution finding
cannot be read as a NON-BLOCKING cleanup.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@briansrls

Copy link
Copy Markdown
Contributor Author

Closing as redundant. This draft was auto-opened on branch session/still-hawk-102-scaffold-exception-reconcile, whose substantive commit (ac53a5b, modeling-discipline scaffold-exception reconcile) already merged to main via #3260 on 2026-05-18. The branch has no effective diff vs main (0 files; only churn merge-commits since). Nothing to implement or flip to ready here.

@briansrls briansrls closed this May 18, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant