Repository navigation
Reconcile ROADMAP generated-projection drift from #5745 (fleet-wide CI unblock) - #5762
Merged
Merged
Conversation
… hand-edit (fleet-wide CI unblock) main's CI is RED on generated_artifact_drift_gate_passes — first_failing_path() reports ROADMAP.md. #5745 (46c33f5) hand-edited ROADMAP.md to add the host-effect orchestration plan pointer line but did NOT update the roadmap authority (.dag), so expected_roadmap_md() drops that line and committed != generated. ROADMAP.md is a GENERATED projection (DESIGN invariant: do not hand-edit, regenerate via main_wet) so every PR merging current main inherits this red — fleet-wide block, no remaining owner (the host-effect/ROADMAP arc session is archived). Content-preserving fix, no model/fold change: the dropped line sits between the "Host-operation band" group header and its first task item, a position a sibling row can't reach while the header is the group's label. So convert that group's label into a section_prose header + add the host-effect pointer as a section_prose sibling (the established `→ [label](path) — desc` idiom, identical shape to the CI-humming pointer row), and empty-label the group (group_blocks already emits header-less for label==""). The host-effect content is preserved verbatim. Verified by execution on the regenerated tree: git diff ROADMAP.md empty after main_wet, first_failing_path() == "", run_generated_artifact_drift_gate == ExitSuccess, roadmap_emit_witnesses still true. Only roadmap_authority.dag changes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This was referenced Jun 24, 2026
briansrls
added a commit
that referenced
this pull request
Jun 24, 2026
This was referenced Jun 24, 2026
Ground the host-effect plan doc into its .dag Plan authority (close #5745's parallel-rep debt)
#5764
Merged
briansrls
added a commit
that referenced
this pull request
Jun 24, 2026
…merge The merge with main brought #5762's structure (Host-operation band label as a section_prose, host-effect pointer below it, empty group label). Conflict was resolved to main's side; regenerate ROADMAP.md from the .dag so emit == committed and the generated_artifact_drift_gate stays green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gunbai-bot Bot
pushed a commit
that referenced
this pull request
Jun 24, 2026
The dashboard auto-committed a stale ROADMAP.md during an interim main_wet regen that ran against the pre-#5762 roadmap_authority (missing #5745's host-effect-orchestration line). Re-merging origin/main picked up #5762's authority reconcile; regenerating restores the canonical projection so generated_artifact_drift_gate passes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
briansrls
added a commit
that referenced
this pull request
Jun 24, 2026
…ENT — replace the global parse-error/unresolved-import short-circuit (front_end_sources collapses graph:None on ANY module's parse error, v1_compiler_compile.rs:2276) with PER-MODULE fail-closed (typed+located diagnostic excludes t (#5760) * WIP: Substrate grounding: make front_end_sources whole-tree-resolve RESILIENT * WIP: Substrate grounding: make front_end_sources whole-tree-resolve RESILIENT * WIP: Substrate grounding: make front_end_sources whole-tree-resolve RESILIENT * WIP: Substrate grounding: make front_end_sources whole-tree-resolve RESILIENT * Regenerate ROADMAP.md projection post-#5762 merge (drift-gate green) The dashboard auto-committed a stale ROADMAP.md during an interim main_wet regen that ran against the pre-#5762 roadmap_authority (missing #5745's host-effect-orchestration line). Re-merging origin/main picked up #5762's authority reconcile; regenerating restores the canonical projection so generated_artifact_drift_gate passes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansearls1@gmail.com> Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
briansrls
added a commit
that referenced
this pull request
Jun 24, 2026
…s + DeclarationRef→NodeRef grounding) (#5754) * WIP: clean up comments as strings * WIP: clean up comments as strings * WIP: clean up comments as strings * WIP: clean up comments as strings * WIP: clean up comments as strings * Model #5745 host-effect-orchestration ROADMAP pointer in the .dag authority #5745 hand-added the `→ [host-effect orchestration plan]` pointer to the generated ROADMAP.md without modeling it in roadmap_authority.dag, so the generated_artifact_drift_gate failed fleet-wide (emit(.dag) lacked the line the committed .md carried) — main has been RED since that merge. An earlier auto-committed regen on this branch "resolved" the drift by deleting the line, which would in turn orphan docs/plans/host-effect-orchestration.md. Fix the authority, not the projection: add the pointer as a section_prose element immediately before the Host-operation band group (its text reads "G1–G3 below are its consumers", which now directly follow it), then regen ROADMAP.md from the .dag. emit(.dag) == committed .md → drift gate green and the host-effect plan doc keeps a modeled inbound link. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * Regen ROADMAP.md to match resolved roadmap_authority.dag after #5762 merge The merge with main brought #5762's structure (Host-operation band label as a section_prose, host-effect pointer below it, empty group label). Conflict was resolved to main's side; regenerate ROADMAP.md from the .dag so emit == committed and the generated_artifact_drift_gate stays green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Brian Searls <briansrls@gunb.ai> Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fleet-wide block
main's own CI is RED on
generated_artifact_drift_gate_passes(verified: recent main runs allfailurefrom #5745 onward).first_failing_path()on clean main reportsROADMAP.md. Every open PR merging current main inherits this red.Root cause
#5745 (commit
46c33f524e, "Host-effect orchestration") hand-added a→ [host-effect orchestration plan](docs/plans/host-effect-orchestration.md)pointer line toROADMAP.mdwithout updating the roadmap authority (dsl/gunbc/roadmap_authority.dag). ROADMAP.md is a generated projection —expected_roadmap_md()doesn't produce that line, so committed ≠ generated → drift. Hand-editing generated artifacts is the exact thing the gate's own failure message forbids.(
roadmap_emit_witnessesis green because it tests emit well-formedness, not byte-equality with the committed file — the authoritative byte-check lives only ingenerated_artifact_gate. So this drift hid from the roadmap-specific witness.)Fix — content-preserving, no model change
The dropped line sits between the "Host-operation band" group header and its first task item — a position a sibling row can't reach while the header is the group's
label. So:labelinto asection_proseheader row,section_prosesibling (the established→ [label](path) — descidiom, byte-identical shape to the existing CI-humming pointer row), andgroup_blocksalready emits header-less whenlabel == "").The host-effect content is preserved verbatim. No edit to
roadmap_model/roadmap_emit(the fold) — authority data only, using existing primitives.Verified by execution (regenerated tree)
git diff ROADMAP.mdempty aftermain_wetfirst_failing_path()==""run_generated_artifact_drift_gate==ExitSuccessroadmap_emit_witnessesstilltruedsl/gunbc/roadmap_authority.dagchanges🤖 Generated with Claude Code