fix(tooling): require the plumbing-skip's sole ctor be reachable as the helper's result

.github/workflows/ci.yml

@@ -11,7 +11,7 @@ jobs:
     permissions:
       contents: read
     strategy:
-      # #848: don't let one failed combo cancel the other 11 — a single
+      # #848: don't let one failed combo cancel the other 12 — a single
       # flaky test cancelled two otherwise-green matrices on PR #846.
       # Full per-run results also make real cross-platform breakage
       # (one OS red, others green) legible at a glance.
@@ -26,6 +26,11 @@ jobs:
         # (Tahoe) is current and dominant (~75%).
         os: [ubuntu-latest, macos-15, macos-26, windows-latest]
         python-version: ["3.11", "3.12", "3.13"]
+        include:
+          # #702: real end-to-end Linux aarch64 coverage — platform
+          # coverage, not python-version coverage, so one job only.
+          - os: ubuntu-24.04-arm
+            python-version: "3.12"
 
     name: test (${{ matrix.os }}, ${{ matrix.python-version }})${{ matrix.python-version == '3.12' && matrix.os == 'ubuntu-latest' && ' + coverage' || '' }}
     runs-on: ${{ matrix.os }}

.github/workflows/limitations-sync.yml

@@ -0,0 +1,102 @@
+name: Limitations sync (issue states)
+
+# Checks that spec/SKILL.md/README limitation tables agree with the live
+# state of the GitHub issues they cite (#852) — e.g. a row still listing a
+# limitation whose tracking issue has since been closed. This is a
+# visibility signal on the Actions tab, not a merge gate: it is
+# deliberately NOT added to required branch-protection status checks,
+# since issue state can drift independently of any particular PR.
+
+on:
+  schedule:
+    - cron: '0 7 * * 1'   # Monday 07:00 UTC — offset from nightly-stress's 06:00 daily slot
+  workflow_dispatch:
+
+# A manual workflow_dispatch run right around the Monday cron shouldn't race
+# a scheduled run against the same live issue-state data; queue instead of
+# cancelling either one.
+concurrency:
+  group: limitations-sync
+  cancel-in-progress: false
+
+jobs:
+  check-states:
+    permissions:
+      contents: read
+      issues: write   # read for the state check; write for the on-failure tracking issue
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - uses: actions/checkout@v7
+        with:
+          persist-credentials: false
+
+      - name: Set up Python 3.12
+        uses: actions/setup-python@v6
+        with:
+          python-version: "3.12"
+
+      - name: Check limitation tables against live issue states
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: python scripts/check_limitations_sync.py --check-states
+
+      # Failure routing, mirroring nightly-stress.yml: scheduled-run
+      # failure emails go to the workflow author, not the maintainer,
+      # and nobody watches the Actions tab — so a failure (drift found,
+      # OR the state check itself couldn't run) files/updates a
+      # labelled tracking issue instead.  Open tracking issues stay
+      # open until manually closed, by design.
+      - name: Open or update tracking issue on cron failure
+        if: failure() && github.event_name == 'schedule'
+        uses: actions/github-script@v9
+        env:
+          RUN_URL: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
+        with:
+          script: |
+            const LABEL = "limitations-drift";
+            const TITLE = "Limitations sync failure on main (tracking)";
+            const runUrl = process.env.RUN_URL;
+            const commentBody = [
+              `The scheduled limitations-sync run failed.`,
+              ``,
+              `**Run logs**: ${runUrl}`,
+              ``,
+              `Either a limitation table cites an issue that has since closed (fix: retire the row per the no-strikethroughs convention), or the state check itself could not run (gh auth / rate limit — the check fails loud rather than passing vacuously).`,
+            ].join("\n");
+
+            // Ensure the label exists (idempotent).
+            try {
+              await github.rest.issues.getLabel({
+                owner: context.repo.owner, repo: context.repo.repo, name: LABEL,
+              });
+            } catch (e) {
+              if (e.status === 404) {
+                await github.rest.issues.createLabel({
+                  owner: context.repo.owner, repo: context.repo.repo,
+                  name: LABEL, color: "d4a72c",
+                  description: "Scheduled limitations-sync failure (#852)",
+                });
+              } else { throw e; }
+            }
+
+            // Find an existing open tracking issue (filter out PRs that
+            // might carry the label — same defensive filter as
+            // nightly-stress.yml).
+            const { data: candidates } = await github.rest.issues.listForRepo({
+              owner: context.repo.owner, repo: context.repo.repo,
+              labels: LABEL, state: "open", per_page: 10,
+            });
+            const existing = candidates.filter(item => !item.pull_request);
+
+            if (existing.length > 0) {
+              await github.rest.issues.createComment({
+                owner: context.repo.owner, repo: context.repo.repo,
+                issue_number: existing[0].number, body: commentBody,
+              });
+            } else {
+              await github.rest.issues.create({
+                owner: context.repo.owner, repo: context.repo.repo,
+                title: TITLE, body: commentBody, labels: [LABEL],
+              });
+            }

AGENTS.md

@@ -8,7 +8,7 @@ Read `SKILL.md` for the full language reference. It covers syntax, slot referenc
 
 ### Conformance programs as reference
 
-The conformance suite in `tests/conformance/` contains 143 small, self-contained programs — often one per language feature — that serve as minimal working examples (the exception is `ch07_cross_module_contracts.vera`, which imports its `ch07_cross_module_contracts_lib.vera` companion to exercise cross-module contracts). Each positive program must pass its declared verification level (see `manifest.json` for mappings: `parse`, `check`, `verify`, or `run`); the seven negative fixtures (`ch02_generic_over_unit_rejected`, `ch05_apply_fn_arity`, `ch08_circular_import`, `ch08_visibility_private`, `ch09_builtin_redefinition`, `ch09_ord_adt_rejected`, `ch09_eq_non_derivable_rejected`) instead must *fail* `check` with the E-code in their `expected_error` field. When you need to see how a specific construct works (e.g. effect handlers, match expressions, closures), check the corresponding conformance program before reading the spec.
+The conformance suite in `tests/conformance/` contains 143 small, self-contained programs — often one per language feature — that serve as minimal working examples (the exceptions are `ch07_cross_module_contracts.vera` and `ch08_cross_module_generic.vera`, which import their `ch07_cross_module_contracts_lib.vera` companion to exercise cross-module contracts). Each positive program must pass its declared verification level (see `manifest.json` for mappings: `parse`, `check`, `verify`, or `run`); the seven negative fixtures (`ch02_generic_over_unit_rejected`, `ch05_apply_fn_arity`, `ch08_circular_import`, `ch08_visibility_private`, `ch09_builtin_redefinition`, `ch09_ord_adt_rejected`, `ch09_eq_non_derivable_rejected`) instead must *fail* `check` with the E-code in their `expected_error` field. When you need to see how a specific construct works (e.g. effect handlers, match expressions, closures), check the corresponding conformance program before reading the spec.
 
 ### Workflow
 
@@ -150,10 +150,10 @@ Read `vera/README.md` for architecture docs, module map, and design patterns.
 ### Pipeline
 
 ```
-source -> parse (parser.py) -> transform (transform.py) -> typecheck (checker.py) -> verify (verifier.py) -> compile (codegen/ + wasm/) -> execute (wasmtime or browser/runtime.mjs)
+source -> parse (parser.py) -> transform (transform.py) -> resolve (resolver.py) -> typecheck (checker.py) -> verify (verifier.py) -> compile (codegen/ + wasm/) -> execute (wasmtime or browser/runtime.mjs)
 ```
 
-Each stage is a module with a single public API function (`parse_file`, `transform`, `typecheck`, `verify`, `compile`, `execute`, `test`) and is independently testable.
+Each stage is a module with a single public API function (`parse_file`, `transform`, `resolve_imports`, `typecheck`, `verify`, `compile`, `execute`, `test`) and is independently testable.
 
 ### Key modules