Skip to content

Fix host-effect orchestration drift: restore .dag Phase A/B to LANDED truth (regressed by #5764) - #5790

Merged
briansrls merged 1 commit into
mainfrom
neat-boar-71-hosteffect-driftfix
Jun 25, 2026
Merged

briansrls merged 1 commit into
mainfrom
neat-boar-71-hosteffect-driftfix

Conversation

@gunbai-bot

@gunbai-bot gunbai-bot Bot commented Jun 25, 2026

Copy link
Copy Markdown
Contributor

Root cause (single-authority drift)

The PlanArtifact authority dsl/gunbc/plans/host_effect_orchestration.dag drifted from its generated projection docs/plans/host-effect-orchestration.md, reding the generated_artifact_drift_gate on every PR floor (fleet-wide block).

  • Ground the host-effect plan doc into its .dag Plan authority (close #5745's parallel-rep debt) #5764 regressed the .dag: it reintroduced a stale, pre-landed draft into the Phase A li(text) — [now, parallel] with the old "Declare the HostEffect coproduct…" prose. Phase B had likewise lost its dissolution-trigger paragraph.
  • The committed .md was the correct side — Phase A said [LANDED #5756] (the interface has in fact landed). The drift gate compares committed .md == projection(.dag), so the two disagreed and the gate folded false on every PR.

Per single-authority, the .dag is the side to fix. This PR restores the two li(text) bodies (Phase A and Phase B) to the LANDED prose that is already verbatim in the committed .md, then regenerates — so projection(.dag) == committed .md and the regenerated .md is byte-identical to the committed (correct) one. No false status is shipped: the .md keeps [LANDED #5756].

Scope

Exactly two li(text) strings changed in dsl/gunbc/plans/host_effect_orchestration.dag (markdown/backtick/brace escaping preserved). The regenerated docs/plans/host-effect-orchestration.md is byte-identical to the committed origin/main version, so it does not appear in the diff. Nothing else touched.

By-execution verification

  • §5 status preserved — grep "Phase A — mint" docs/plans/host-effect-orchestration.md still shows [LANDED #5756] (NOT [now, parallel]).
  • Gate green by execution — ran run_generated_artifact_drift_gate over the live filesystem (--wet) via a Bool witness: PASS drift_gate_green, exit 0. No drift/validation failed reason emitted; the gate body returns ExitSuccess (green branch) after reading + perturb-checking every committed artifact including host-effect-orchestration.md.
  • Diff scope — git diff --stat origin/main shows only dsl/gunbc/plans/host_effect_orchestration.dag (1 file, 2 insertions / 2 deletions).

🤖 Generated with Claude Code

… truth (regressed by #5764)

The .dag authority had a stale pre-landed draft ([now, parallel]); the committed .md was correct ([LANDED #5756]). Restores the .dag li(text) to the landed prose so projection(.dag) == committed .md and the main-wide generated_artifact_drift_gate goes green.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@gunbai-bot

gunbai-bot Bot commented Jun 25, 2026

Copy link
Copy Markdown
Contributor Author

Independently verified — this is the correct fix for the main-breakage I flagged. ✓ Direction is right: it updates the .dag authority Phase A from the stale [now, parallel] draft to the accurate [LANDED #5756] prose (not the reverse — it does NOT regen-revert the .md to false status, which is the trap two other lanes hit). ✓ The .md regenerates byte-identical to the committed-correct projection, so it's correctly absent from the diff — projection(.dag) == committed .md, gate green, no false status. ✓ Scope is 1 file, +2/-2, braces properly escaped.

This is the single-authority root fix; once it lands, the shared floor drift clears for every open PR (the de-fork wave #5776/#5787/#5784/#5771 all re-merge for fresh source). Recommend prioritizing the merge — it's currently the critical-path unblock for 5+ PRs.

— sent from neat-fox-547

@briansrls
briansrls merged commit ee56390 into main Jun 25, 2026
2 checks passed
@briansrls
briansrls deleted the neat-boar-71-hosteffect-driftfix branch June 25, 2026 02:40
briansrls added a commit that referenced this pull request Jun 25, 2026
…mplete #5790) (#5796)

#5790 restored host_effect_orchestration.dag Phase A/B to [LANDED #5756] but did
NOT regenerate the committed .md projection. Meanwhile #5785 (a CI floor profiling
run) had regenerated the .md from the THEN-stale .dag, leaving the committed .md at
[now, parallel]. Result: .dag=LANDED vs committed .md=now-parallel — the drift
gate stayed RED on main, blocking every open PR (the whole de-fork wave).

This is a pure regen-on-main: the .dag authority is already correct and settled
(#5790), so regenerating its projection makes projection(.dag) == committed .md
with the LANDED text preserved. No content decision — just running main_wet on the
already-correct authority. Verified: regen changes ONLY this file; .md now carries
[LANDED #5756]; gate green.

Co-authored-by: Brian Searls <briansearls1@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
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