Repository navigation
Structural-quadratic walls DELIVERABLE 1: complexity-mechanism coverage audit + execution receipt - #6162
Merged
Conversation
…ge audit + execution receipt Census-first, per the brief: inventory every complexity mechanism (gate / lens / budget / ratchet) and locate each on the expressibility-frontier trichotomy (wall / lens-residue / review) BEFORE proposing any new wall. New modeled plan gunbc.plans.structural_quadratic_wall_coverage_audit (projected to docs/plans/structural-quadratic-wall-coverage-audit.md, registered in plan_registry_batch_g, linked from the frontier plan so it is not an orphan doc). Contents: - Execution receipt: ran the emit-only complexity_linearity_audit bin over main@6e881a11 (1132 files, 10212 fns, 428 syntactic findings; cost proxy fired on 22 sites; wildcard arms 406). - 13-row inventory table classifying every cost/complexity/budget mechanism by class, floor-enrollment, and structural-vs-resource. - Verdict: the decidable structural-quadratic WALL already exists (src/v2/lens/cost.dag WallNow + termination DescentEvidence + the memory actuator), but is roster-bound; whole-corpus reach is region-(2) residue blocked on the same decl_facts(roots) #5966 / gunbc#5364 reflection that shelves ROADMAP §3 `3-gates-whole`. Recommendation: build no new oracle; the only on-dial increment is a post-fix reintroduction fence for the resolver-pathology quadratic, sequenced AFTER the sibling lanes land. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gunbai-bot
Bot
changed the title
Relocate the coverage-audit plan to dag/gunbc/plans/; keep the pre-push freshness entry for docs/plans/structural-quadratic-wall-coverage-audit.md alongside main's dag/** roster. Regen follows to reconcile generated docs.
…grounding) main renamed the substrate tree dsl/->dag/; the coverage-audit inventory column still cited dsl/std/*, dsl/extdeps/*, dsl/gunbc/* paths that no longer exist. Retargeted all 10 to dag/ and regenerated the mirror markdown (main_wet). Addresses cursor/composer-2.5 REQUEST_CHANGES on #6162. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
briansrls
added a commit
that referenced
this pull request
Jul 4, 2026
…resentation prevention wall (#6192) * WIP: STRUCTURAL-QUADRATIC WALLS (not a ratchet): make the OBVIOUS, DECIDABLE * Structural-quadratic walls DELIVERABLE 1: complexity-mechanism coverage audit + execution receipt Census-first, per the brief: inventory every complexity mechanism (gate / lens / budget / ratchet) and locate each on the expressibility-frontier trichotomy (wall / lens-residue / review) BEFORE proposing any new wall. New modeled plan gunbc.plans.structural_quadratic_wall_coverage_audit (projected to docs/plans/structural-quadratic-wall-coverage-audit.md, registered in plan_registry_batch_g, linked from the frontier plan so it is not an orphan doc). Contents: - Execution receipt: ran the emit-only complexity_linearity_audit bin over main@6e881a11 (1132 files, 10212 fns, 428 syntactic findings; cost proxy fired on 22 sites; wildcard arms 406). - 13-row inventory table classifying every cost/complexity/budget mechanism by class, floor-enrollment, and structural-vs-resource. - Verdict: the decidable structural-quadratic WALL already exists (src/v2/lens/cost.dag WallNow + termination DescentEvidence + the memory actuator), but is roster-bound; whole-corpus reach is region-(2) residue blocked on the same decl_facts(roots) #5966 / gunbc#5364 reflection that shelves ROADMAP §3 `3-gates-whole`. Recommendation: build no new oracle; the only on-dial increment is a post-fix reintroduction fence for the resolver-pathology quadratic, sequenced AFTER the sibling lanes land. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * Merge origin/main (dsl/->dag/ rename) into session/silent-ferret-137 Relocate the coverage-audit plan to dag/gunbc/plans/; keep the pre-push freshness entry for docs/plans/structural-quadratic-wall-coverage-audit.md alongside main's dag/** roster. Regen follows to reconcile generated docs. * Review fix (#6162): retarget audit inventory carriers dsl/->dag/ (§3 grounding) main renamed the substrate tree dsl/->dag/; the coverage-audit inventory column still cited dsl/std/*, dsl/extdeps/*, dsl/gunbc/* paths that no longer exist. Retargeted all 10 to dag/ and regenerated the mirror markdown (main_wet). Addresses cursor/composer-2.5 REQUEST_CHANGES on #6162. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * Complexity gating (i): completeness-gated import-closure parallel-representation prevention wall First real enforced wall from the complexity-gating lane (Deliverable 1 audit: detection real, enforcement zero). A change-time / §2-§4 STRUCTURAL wall: a module's DECLARED consumed-input closure must equal its DERIVED import closure (import_closure_live). It FOLDS the whole declared_consumed_input_closures registry, so any future declared closure is auto-covered — closing the "only the one conformance closure is hand-witnessed" gap. Forward-looking ask-once-compile-forever prevention; registry is 1 clean entry (0 drift) so it lands clean, no self-poison. Adds entry_path to ConsumedInputClosure (additive; existing consumers read consumed_dag_paths/unit only; conformance witness regression PASS). Green-by-execution on the v1 floor engine (claim_batch), 3 witnesses: clean (0 drift), red control (drifted closure detected), completeness (injected drift into the fold auto-caught). Enforcement maturity (honest): realized as a floor witness (fail-closed — a red witness fails the floor). A corpus-level registry-fold check does not fit always_required_lenses (per-compiled-node CompileLens); the floor witness is the correct fail-closed realization. Because main ci.yml cancels at the 30-min compile-clean timeout before batch-2 witnesses run, this wall — like every batch-2 witness — enforces on the local/scheduled floor run, not yet on main CI; full CI enforcement follows the compile-clean-serial lane. Runtime-quadratic walls remain blocked on the v2 body-lowering milestone (real fn bodies do not lower to inspectable Loop nodes); dispatched separately. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * WIP: STRUCTURAL-QUADRATIC WALLS (not a ratchet): make the OBVIOUS, DECIDABLE * Revert stray auto-committed audit-row edit (restore .dag<->.md consistency) The v2 DescentEvidence-fork audit rows were mid-edit when the auto-committer captured them into 105a18c without the paired plan-doc regen (which routes to BuildBuddy remote and cannot see a working-tree edit), leaving the committed .dag drifted from its generated .md projection = a drift-gate self-poison. The fold is deferred to fierce-bee's body-lowering Stage-1, where the DescentEvidence fork is consolidated (so the row's final content is determined there) and the plan-doc regen is already part of that lane's workflow. 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.
DELIVERABLE 1 (census-first) — complexity-mechanism coverage audit + execution receipt
Per the brief, the census comes FIRST: inventory every complexity mechanism and locate each on the expressibility-frontier trichotomy (①wall / ②lens-residue / ③review) before proposing any new wall — so the wall decision is determined by the census, not searched.
Deliverable: new modeled plan
gunbc.plans.structural_quadratic_wall_coverage_audit, projected todocs/plans/structural-quadratic-wall-coverage-audit.md, registered inplan_registry_batch_g, linked from the frontier plan (not an orphan doc).Execution receipt (main @ 6e881a1)
Ran the emit-only
complexity_linearity_auditbin: 1132 files, 10212 fns, 428 syntactic findings; cost proxysyntactic_high_match_fanout= 22 sites; wildcard arms 406. This whole-corpus mechanism is a syntactic proxy, NOT the real asymptotic engine.Verdict
src/v2/lens/cost.dag(WallNow, foldsSymbolicCost, detectsPoly(2)structurally) +termination.dagDescentEvidence+ therealization_widthmemory actuator — but is roster-bound (never walks the 1132-file corpus).decl_facts(roots)Whole-tree EXTRACT for model-grounding lens: concept_decl_facts(roots), name index, layer exclusion, worklist schema #5966 / gunbc#5364 reflection that shelves ROADMAP §33-gates-whole.AdvisoryLens/RatchetForever.Recommendation
Build no new complexity oracle or cost-inference algebra (both brief non-goals; a second engine forks the cost authority). The only on-dial increment is a post-fix reintroduction fence for the resolver-pathology quadratic shape, sequenced after the sibling lanes (bold-ibex-109 compositional
build_type_env, representation-minimization) land — the one candidate that could displace a measured cost on a pillar path (ROADMAP §4 un-shelve bar).Floor-safe: modeled plan + regenerated projection (
main_wet), inbound-linked, no load-bearing files touched.🤖 Generated with Claude Code