feat(wave5): per-language testing standard v2.0.0 + guide template + AffineScript

The estate's only per-language testing depth was a single Julia guide from 2024
(no MUST/SHOULD, Rust+Julia only) plus a byte-identical duplicate snapshot. The
user flagged this: replicate the pattern properly instead of one stale generic.

- standards/language-testing-standards.md -> v2.0.0: RFC-2119 conformance
  requirements R1–R9 (unit/format/lint/coverage/property/bench/audit/contract/
  proof) mapped to the CRG test taxonomy, an anti-theatre rule (no
  continue-on-error on a MUST check; coverage reported-not-asserted), and the
  required per-language guide set. Rust/SPARK + Julia mapped to R1–R9.
- standards/templates/language-testing-guide-TEMPLATE.md: the skeleton every
  per-language guide follows — a requirement-mapping table (tool or visible
  `none`), tools, CI pipeline, best practices, and a mandatory honest
  "Known gaps" section.
- standards/affinescript-testing-guide.md: the estate's primary language, which
  had ZERO testing standard. Authored honestly — most SHOULD rows are tracked
  gaps (no formatter/coverage/fuzz/bench yet) and R3 notes that
  affinescript-verify.yml is currently advisory. Canonical SSOT migrates to
  hyperpolymath/affinescript prospectively (Wave-6 charter).
- scripts/check-language-guide.sh + Justfile language-guides-check, wired into
  `just validate` as a hard gate: a guide missing a required section (e.g.
  "Known gaps"), the SPDX header, or the R1..R9 mapping fails loudly.
- Deleted the byte-identical language-testing-standards-v1.0.0-2024-04-14.md.
- scripts/tests/wave5-language-guides-test.sh (7/7): lint accepts a conformant
  guide and rejects incomplete ones; the standard is v2.0.0/RFC-2119.

Zig / Elixir+Gleam / Idris2+Agda guides are Wave-6 charters. Licence rows
manual-only (flag-only).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0114ps6mY5jAH4Sz

Author: claude

Justfile

@@ -85,12 +85,18 @@ dyadt-conformance:
 dyadt-test:
     @bash scripts/tests/wave4-dyadt-test.sh
 
+# Structural lint for per-language testing guides (required sections + R1..R9)
+language-guides-check:
+    @bash scripts/check-language-guide.sh
+
 # Aggregate compliance gate: registry drift is the HARD gate (registry-check,
 # a hard dep). The RSR self-audit is INFORMATIONAL — a monorepo is not expected
 # to score Gold — but a *broken* audit (exit 4 / unexpected) must fail loudly
 # rather than pass silently under a blanket `|| true` (Wave-0 false-green fix).
 validate: registry-check
     @echo "=== validate: registry drift (HARD GATE) — passed as a dependency above ==="
+    @echo "=== validate: per-language testing guides (structural, HARD GATE) ==="
+    @bash scripts/check-language-guide.sh
     @echo "=== validate: RSR self-audit (INFORMATIONAL grade; errors fail loudly) ==="
     @bash scripts/rsr-selfaudit.sh .
     @echo "=== validate: done ==="

scripts/check-language-guide.sh

@@ -0,0 +1,55 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: MPL-2.0
+#
+# check-language-guide.sh — structural lint for per-language testing guides.
+#
+# Every guide built from templates/language-testing-guide-TEMPLATE.md MUST carry
+# the required sections and the R1–R9 requirement-mapping table. A guide that
+# silently omits a section (e.g. "Known gaps") is a false-completeness hole —
+# this fails loudly instead.
+#
+# Usage: check-language-guide.sh [guide.md ...]
+#   With no args, checks every standards/*-testing-guide.md.
+# Exit: 0 all valid · 1 a guide is missing a required section
+
+set -uo pipefail
+
+ROOT="$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
+
+REQUIRED_SECTIONS=(
+  "## Requirement mapping"
+  "## Tools"
+  "## Recommended CI pipeline"
+  "## Best practices"
+  "## Known gaps"
+  "## Resources"
+)
+
+check_one() { # file
+  local f="$1" rc=0 sec
+  if [ ! -f "$f" ]; then echo "  ❌ $f: not found"; return 1; fi
+  for sec in "${REQUIRED_SECTIONS[@]}"; do
+    grep -Fqx "$sec" "$f" || { echo "  ❌ $(basename "$f"): missing section '$sec'"; rc=1; }
+  done
+  # The requirement mapping MUST reference the R1..R9 rows (at least R1 and R9).
+  grep -Eq '\bR1\b' "$f" && grep -Eq '\bR9\b' "$f" || { echo "  ❌ $(basename "$f"): requirement mapping does not reference R1..R9"; rc=1; }
+  # A SPDX header is required.
+  head -3 "$f" | grep -q 'SPDX-License-Identifier' || { echo "  ❌ $(basename "$f"): missing SPDX header"; rc=1; }
+  [ "$rc" -eq 0 ] && echo "  ✅ $(basename "$f")"
+  return $rc
+}
+
+if [ "$#" -gt 0 ]; then
+  files=("$@")
+else
+  mapfile -t files < <(ls "$ROOT"/standards/*-testing-guide.md 2>/dev/null)
+fi
+
+if [ "${#files[@]}" -eq 0 ]; then
+  echo "no language testing guides found (standards/*-testing-guide.md)"; exit 0
+fi
+
+rc=0
+echo "Language testing guides:"
+for f in "${files[@]}"; do check_one "$f" || rc=1; done
+exit $rc

scripts/tests/wave5-language-guides-test.sh

@@ -0,0 +1,41 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: MPL-2.0
+set -uo pipefail
+#
+# Wave-5 regression: the per-language testing guide lint must accept a
+# template-conformant guide and reject one missing a required section.
+
+ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+CHK="$ROOT/scripts/check-language-guide.sh"
+TMP="$(mktemp -d)"; trap 'rm -rf "$TMP"' EXIT
+
+pass=0 fail=0
+ok()  { echo "  ✅ $1"; pass=$((pass + 1)); }
+bad() { echo "  ❌ $1"; fail=$((fail + 1)); }
+
+echo "== real guides pass =="
+bash "$CHK" >/dev/null 2>&1 && ok "estate guides pass structural lint" || bad "estate guides failed lint"
+bash "$CHK" "$ROOT/standards/affinescript-testing-guide.md" >/dev/null 2>&1 && ok "affinescript guide valid" || bad "affinescript guide invalid"
+
+echo "== rejects incomplete guides =="
+# missing a required section
+g="$TMP/foo-testing-guide.md"
+printf '<!-- SPDX-License-Identifier: CC-BY-SA-4.0 -->\n# Foo\n## Requirement mapping\nR1 ... R9 ...\n## Tools\n## Recommended CI pipeline\n## Best practices\n## Resources\n' > "$g"
+bash "$CHK" "$g" >/dev/null 2>&1 && bad "missing 'Known gaps' not caught" || ok "missing section rejected"
+# missing SPDX header
+g2="$TMP/bar-testing-guide.md"
+printf '# Bar\n## Requirement mapping\nR1 R9\n## Tools\n## Recommended CI pipeline\n## Best practices\n## Known gaps\n## Resources\n' > "$g2"
+bash "$CHK" "$g2" >/dev/null 2>&1 && bad "missing SPDX not caught" || ok "missing SPDX rejected"
+# missing R1..R9 reference
+g3="$TMP/baz-testing-guide.md"
+printf '<!-- SPDX-License-Identifier: CC-BY-SA-4.0 -->\n# Baz\n## Requirement mapping\nno numbers here\n## Tools\n## Recommended CI pipeline\n## Best practices\n## Known gaps\n## Resources\n' > "$g3"
+bash "$CHK" "$g3" >/dev/null 2>&1 && bad "missing R1..R9 not caught" || ok "missing R1..R9 rejected"
+
+echo "== the stale duplicate snapshot is gone =="
+[ ! -f "$ROOT/standards/language-testing-standards-v1.0.0-2024-04-14.md" ] && ok "duplicate snapshot removed" || bad "duplicate snapshot still present"
+echo "== the standard is v2.0.0 with RFC-2119 =="
+grep -q 'Version:\*\* 2.0.0' "$ROOT/standards/language-testing-standards.md" && grep -qi 'RFC-2119' "$ROOT/standards/language-testing-standards.md" && ok "standard refreshed to v2.0.0 RFC-2119" || bad "standard not refreshed"
+
+echo
+echo "Wave-5 language-guides regression: $pass passed, $fail failed"
+[ "$fail" -eq 0 ]

standards/affinescript-testing-guide.md

@@ -0,0 +1,109 @@
+<!-- SPDX-License-Identifier: CC-BY-SA-4.0 -->
+# AffineScript Testing Tools Guide
+
+**Version:** 1.0.0
+**Date:** 2026-07-03
+**Status:** Active (baseline — honest by construction)
+**Parent standard:** `language-testing-standards.md` (R1–R9)
+
+AffineScript is the estate's primary application language (RS/TS/JS →
+AffineScript → typed-wasm; affine/linear types, OCaml-based compiler). This
+guide is the estate's current best statement of its testing story. AffineScript's
+tooling is young, so several rows below are honest **gaps**, not omissions — a
+gap here is a tracked piece of work, and this guide names it rather than
+pretending coverage exists.
+
+**Canonical SSOT (prospective):** the authoritative home for this guide will be
+`hyperpolymath/affinescript` (`spec/` or `docs/testing.adoc`). Until that lands,
+this repo carries it; the migration is a Wave-6 charter. Do not let the two
+diverge — when the affinescript-repo version ships, this becomes a pointer.
+
+## Requirement mapping
+
+| # | Requirement | Level | Tool | CI invocation | Status |
+|---|---|---|---|---|---|
+| R1 | Unit test runner | MUST | `affinescript-deno-test` bootstrap runner | `deno task test` (in the AS repo) | partial — bootstrap shim, self-hosting pending |
+| R2 | Formatter (check mode) | MUST | `affinescript fmt` (compiler subcommand) | `affinescript fmt --check` | gap — formatter not yet shipped |
+| R3 | Linter / static analysis | MUST | compiler diagnostics + `affinescript-verify.yml` | `affinescript compile --check` | partial — the type system IS the primary static check; a dedicated linter is a gap |
+| R4 | Coverage | SHOULD | `none` | — | gap |
+| R5 | Property-based / fuzz | SHOULD | `none` | — | gap (parser/lowering are the priority targets) |
+| R6 | Benchmark | SHOULD | wasm bench harness | — | gap |
+| R7 | Security / dependency audit | MUST | Deno (`deno.json` import audit) | `deno task audit` | partial — Deno-managed deps; no AS-native audit |
+| R8 | Contract / pre-post | MAY | affine/linear types (compile-time) | compiler | partial — linearity is a compile-time contract |
+| R9 | Proof check | MUST\* | Idris2 ABI proofs (for proven-backed modules) | ECHIDNA proof gate | partial — applies to modules using the `proven` library |
+
+`MUST*` = R7 applies (Deno ecosystem); R9 applies only to AS modules that call
+proven/Idris2-verified code.
+
+## Tools
+
+### AffineScript compiler (`affinescript`) — R2, R3, R8
+- **Purpose:** the type checker is the primary correctness gate. Affine/linear
+  types reject use-after-move and aliasing at compile time — that is R8
+  (contract) discharged by construction, and much of R3 (static analysis).
+- **Usage:** `affinescript compile <file>.affine` (type-checks + lowers to
+  typed-wasm); `affinescript compile --check` for check-only.
+- **CI:** `.github/workflows/affinescript-verify.yml` clones + builds the
+  compiler and runs verification. **Note:** that job is currently *advisory*
+  (`continue-on-error`) while the compiler build stabilises — it does not yet
+  gate. Promotion to blocking is the unblock condition for R3.
+
+### affinescript-deno-test — R1
+- **Purpose:** the bootstrap test runner used until AffineScript self-hosts its
+  test framework. TS/JS shim (documented carve-out).
+- **Usage:** `deno task test` in the AS repo.
+- **CI:** runs in the AffineScript repo's CI.
+
+## Recommended CI pipeline
+
+Until the AS-native toolchain matures, the recommended pipeline is:
+
+1. **Type-check (R3/R8, MUST):** build the compiler, `affinescript compile
+   --check` over all `.affine` sources — SHA-pinned, and **blocking once the
+   compiler build is reliably green** (today advisory; see `affinescript-verify.yml`).
+2. **Unit tests (R1, MUST):** `deno task test` via the bootstrap runner.
+3. **Dep audit (R7, MUST):** Deno import audit.
+4. SHOULD rows (coverage, property, bench) are tracked gaps — see below.
+
+No `continue-on-error` on a MUST check once its tool is stable; the current
+advisory status of `affinescript-verify.yml` is itself a tracked gap, not a
+silent pass.
+
+## Best practices
+
+1. Design modules to admit affine/linear typing from the start — the type system
+   is the cheapest test you have.
+2. Prefer compile-time linearity contracts (R8) over runtime assertions where the
+   type system can express the invariant.
+3. For correctness-critical paths, route through `proven`/Idris2 (R9) rather than
+   hand-rolled checks.
+4. Keep `.affine` sources free of TS/JS shims except the documented bootstrap
+   carve-outs.
+
+## Known gaps
+
+Honest inventory (every gap is real work, not an omission):
+
+- **R2 formatter** — no `affinescript fmt` yet. Charter.
+- **R3 dedicated linter** — beyond type diagnostics; and `affinescript-verify.yml`
+  is advisory (`continue-on-error`), so R3 does not yet *gate*. Charter: flip to
+  blocking once the compiler build is reliably green.
+- **R4 coverage** — no wasm coverage tool. Charter.
+- **R5 property/fuzz** — none; parser and canonical-lowering are the priority
+  targets. Charter.
+- **R6 benchmark** — no wasm bench harness. Charter.
+- **R1 self-hosting** — the test runner is a TS/JS bootstrap shim, not AS-native.
+  Unblocks when AffineScript self-hosts the runner.
+
+These gaps are why AffineScript's Toolchain Readiness Grade cannot yet exceed the
+lower bands — which is the honest position, and the reason this guide exists.
+
+## Resources
+
+- `language-testing-standards.md` — the parent R1–R9 standard.
+- `.github/workflows/affinescript-verify.yml` — the current (advisory) CI check.
+- `templates/language-testing-guide-TEMPLATE.md` — the skeleton this follows.
+- SSOT (prospective): `hyperpolymath/affinescript`.
+
+**Maintainers:** @hyperpolymath
+**Last Updated:** 2026-07-03