briansrls · May 15, 2026 · May 15, 2026 · May 15, 2026 · May 15, 2026 · May 15, 2026
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -48,6 +48,7 @@ jobs:
     outputs:
       v2: ${{ steps.detect.outputs.v2 }}
       v3: ${{ steps.detect.outputs.v3 }}
+      v4: ${{ steps.detect.outputs.v4 }}
     steps:
     - name: Checkout
       uses: actions/checkout@v4
@@ -174,6 +175,46 @@ jobs:
       run: |
         set -euo pipefail
         cargo test -p v2-compiler-tests --release bootstrap_fixed_point -- --ignored --exact
+  # v4 bootstrap viability — verifies v2's compiled binary can still parse
+  # every v4 .dag file. v4 stage0 must stay in v2-syntax-compatible subset
+  # (per src/v4/STRUCTURE.md "Bootstrap chain") until v4 self-compiles.
+  # This gate catches any v4 change that introduces syntax v2 can't parse.
+  v4:
+    if: github.event.pull_request.draft != true && needs.affected.outputs.v4 == 'true'
+    needs: [affected]
+    runs-on: ${{ vars.CI_RUNNER || 'ubuntu-latest' }}
+    timeout-minutes: 15
+    steps:
+    - name: Checkout
+      uses: actions/checkout@v4
+      with:
+        fetch-depth: 1
+    - name: Setup Rust
+      uses: actions-rust-lang/setup-rust-toolchain@v1.16.0
+      with:
+        cache: false
+        rustflags: ''
+    - name: Cache Cargo (v4 — shares with v2 since we build v2 binary)
+      uses: actions/cache@v4
+      with:
+        path: |
+          ~/.cargo/registry/index/
+          ~/.cargo/registry/cache/
+          ~/.cargo/git/db/
+          target/
+        key: cargo-v2-${{ runner.os }}-${{ hashFiles('**/Cargo.lock') }}-${{ hashFiles('src/v2/**') }}
+        restore-keys: |
+          cargo-v2-${{ runner.os }}-${{ hashFiles('**/Cargo.lock') }}-
+          cargo-v2-${{ runner.os }}-
+    - name: Build v2 compiler (bootstrap stage minus one)
+      run: cargo build -p v2-compiler --release
+    - name: v2 → v4 bootstrap viability (parse all v4 .dag files via v2)
+      run: |
+        set -euo pipefail
+        rm -rf /tmp/v4-stage1
+        mkdir -p /tmp/v4-stage1
+        target/release/v2-compiler compile --source-root src/v4 --output-dir /tmp/v4-stage1 --target dag
+        echo "Bootstrap viability OK — v2 indexed and resolved all v4 modules."
   # v3 — FROZEN 2026-05-15 pending v4 program. Runs only when src/v3/ or dsl/
   # is affected (or on push-to-main for trunk verification). v3 is no longer
   # the active compiler; v4 program supersedes it (see PR #3147).

diff --git a/ROADMAP.md b/ROADMAP.md
@@ -2,6 +2,8 @@
 
 Single source of truth for project status, active work, and deferred items. Long-form receipts and historical narratives now live under `docs/history/` and `docs/db-history/` so this file can stay operational.
 
+> **v4 supersession (2026-05-15).** The active development phase is **v4** — see [`src/v4/STRUCTURE.md`](src/v4/STRUCTURE.md), [`src/v4/TASKS.md`](src/v4/TASKS.md), [`docs/v4-close-interrogation.md`](docs/v4-close-interrogation.md). v4 combines R3 + R4 into one ship per zero-deferrals discipline; the R1 program below is being superseded by v4's XL task plan. v3 is frozen (CI gated to `src/v3/`-affected only). This roadmap doc retains R1 lane structure as historical context until v4-driven work fully replaces it; lane-by-lane translation lands incrementally.
+
 > Design spec: [docs/v3-spec.md](docs/v3-spec.md)
 > Validation: [docs/v3-validation-experiments.md](docs/v3-validation-experiments.md)
 > Lineage: [docs/design-lineage.md](docs/design-lineage.md)

diff --git a/THESIS.md b/THESIS.md
@@ -2,6 +2,8 @@
 
 This is the parent document. Everything else — ROADMAP, INVARIANTS, MODELING, architecture, and design docs — serves this thesis.
 
+> **v4 supersession (2026-05-15).** The active program is **v4** in [`src/v4/`](src/v4/) — substrate scaffold + R3+R4 combined ship per [`docs/v4-close-interrogation.md`](docs/v4-close-interrogation.md). Every thesis claim below applies to v4; v3 references in the §Self-hosting four facets section (line ~232+) describe the v2→v3 transition that v4 supersedes (v4 = "v2's residual discipline + v3's modeling depth + recursive-flex from day 1"). v4's substrate is the operational instantiation of every claim in this thesis; v3 is frozen pending v4 ship. See [`src/v4/STRUCTURE.md`](src/v4/STRUCTURE.md) for the v4 file tree and [`docs/v4-close-interrogation.md`](docs/v4-close-interrogation.md) §"v4 applicability mapping" for v3→v4 path translation.
+
 ## How this doc is organized
 
 Read this file for the thesis itself and the complete claims list. Extended argumentation now lives under `docs/thesis/`, while `ROADMAP.md`, `INVARIANTS.md`, and `MODELING.md` stay as the live operational companions.

diff --git a/docs/design-pure-bootstrap-zero.md b/docs/design-pure-bootstrap-zero.md
@@ -2,6 +2,8 @@
 
 **Status:** `LIVE` (promoted 2026-04-25 via cascade promotion [PR #782](https://github.com/gunb-ai/gunbc/pull/782)). Supersedes the ≤5-floor framing in [`docs/design-pure-bootstrap.md`](design-pure-bootstrap.md) (now marked SUPERSEDED with explicit retraction) and the 2-3 principled-floor framing introduced via [PR #756](https://github.com/gunb-ai/gunbc/pull/756) (also retracted). This doc is the live authority on the Pure Bootstrap to Zero program.
 
+> **v4 supersession (2026-05-15).** The 0-floor target articulated in this doc now applies to **v4** ([`src/v4/`](../src/v4/)) as the operational instantiation. v2 binary serves as v4's stage minus one (per [`src/v4/STRUCTURE.md`](../src/v4/STRUCTURE.md) "Bootstrap chain"); v4's compiler emits its own Rust trampoline (`bin/main.dag`) to satisfy 0-floor without needing the runtime-resolution choices (shipped binary / runtime crate / rustc-macro) described below. v3 references throughout this doc describe the v2→v3 transition that v4 supersedes; v3 is frozen pending v4 ship.
+
 **Promotion evidence chain (cited in cascade promotion PR body):**
 - D1 audit: PRs #769 + #771 + #775 + #777 + #779 (audit doc with substrate-generation already proven; 23 generated files + 24 REGEN_OUTPUTS entries; 38-type substrate.dag coverage survey)
 - D2 PB-1 brief amendment: PR #770 (non-goals revised under 0-floor)

diff --git a/docs/thesis-claim-coverage.md b/docs/thesis-claim-coverage.md
@@ -197,8 +197,8 @@ This doc is that audit. Every thesis claim in `THESIS.md §"Thesis claims — co
 
 | # | Claim | Disposition | Evidence | Status |
 |---|---|---|---|---|
-| 64 | (R1) Suboptimal-complexity contract violation: function annotated bound errors at compile | R1-closed | T-Demo `impossible_bug_class_suite_r1` includes suboptimal-complexity demo | partial (CostBounded receipt is runner-fail) |
-| 65 | (R1) Idempotency-contract violation: marked `@idempotent` function structure errors | R1-closed | T-Demo impossible-bugs suite (`compose_effects` + `AppendEffect` violation) | green |
+| 64 | (R1) Suboptimal-complexity contract violation: lens-derived complexity exceeds the structural cost bound (consumer/inhabitance) → compile error (no annotation) | R1-closed | T-Demo `impossible_bug_class_suite_r1` includes suboptimal-complexity demo | partial (CostBounded receipt is runner-fail) |
+| 65 | (R1) Idempotency-contract violation: structurally non-idempotent operation in an idempotence-required (algebra-inhabitance) position → compile error (no annotation) | R1-closed | T-Demo impossible-bugs suite (`compose_effects` + `AppendEffect` violation) | green |
 | 66 | (R1) Transport/type-drift: client/server cannot hold different types for same field | R1-closed | T-Demo impossible-bugs suite (`TypeMismatch` multi-target) | green |
 | 67 | (R2+) Nested-optional flatten: `Option<Option<T>>` patterns error at compile | R2-gated | `docs/r2-structure.md` Goal 4; gated on cardinality refinement; closed by PR #890 / `6562081e` | green |
 | 68 | (R2+) Unenumerated effects: operations intrinsically read/write via type signature | R2-gated | `docs/r2-structure.md` Goal 4; `docs/briefs/t-impossiblebugs-unenumerated-effects-design.md` + `docs/briefs/r2-impossible-bugs-unenumerated-effects-worker.md` closed-system structural derivation path | in-flight (implementation worker dispatchable; consumer/lens PR active) |