feat: delete the legacy 17-row eval matrix — the sliced lane is the only paid lane

Every PR paid twice: the hand-enumerated matrix (18 test files, 22.6 min,
~$21 API measured on run 33263204465) ran serialized AHEAD of the strictly
superior sliced lane via 'needs: evals' — 35.5 min wall and ~2x paid spend
for the same diff. 14 of 17 rows carried no tier:, so periodic Opus
benchmarks leaked into every PR (the e2e-plan row alone: 12/12 tests,
21.7 min, $7.28 — the wall-clock bound of ALL of CI).

Parity receipt (static, pre-deletion): the sliced lane's gate census (49
files, derived from the runner itself) strictly contains all 18 matrix test
files, plus 31 files the matrix never ran. Pure deletion — one revert
restores it. The PR comment moved into slices-report (same '## E2E Evals'
upsert marker, now sourced from slice artifacts + carrying the fail-closed
reconciliation verdict). plan-slices loses the needs edge; the dead
workflow-level EVALS_TIER env goes with it.

test/evals-workflow-matrix.test.ts (and its KNOWN_MATRIX_GAPS /
KNOWN_TIER_UNSET burn-down ratchets — retired: the sliced census makes
'every gate file runs' true by construction) is rewritten as
test/evals-workflow-wiring.test.ts: matrix stays deleted, planner/executor/
report tier + slice-count agreement, both surviving lanes on the shared
register-skills composite with its fail-fast verification loop, PR comment
survival. Expected: PR eval wall 35.5 -> ~13 min, per-PR paid spend ~halved.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

TODOS.md

@@ -567,14 +567,15 @@ duration-packed free shards, the sharded paid runner as the CI engine
 coverage contract + gate census, eval-budget timeout tiers, and the
 coverage fill. Remaining, in rough priority order:
 
-- **P1 — Delete the legacy evals.yml matrix after parity.** The sliced lane
-  runs alongside the 18-row matrix (`needs: evals`, so provider concurrency
-  never doubles). After 1-2 PR cycles of parity (compare executed-test sets:
-  intersection strict + the 8 KNOWN_MATRIX_GAPS files as expected additions;
-  stochastic outcomes informational), delete the matrix as a PURE-DELETION
-  commit (one revert restores it), drop the `needs: evals` edge, rewrite
-  test/evals-workflow-matrix.test.ts into a runner-wiring pin, and retire
-  KNOWN_MATRIX_GAPS/KNOWN_TIER_UNSET wholesale. Effort S.
+- **DONE (v1.76 test-infra wave) — Delete the legacy evals.yml matrix after
+  parity.** Deleted as a pure-deletion commit (one revert restores it) after
+  a static parity receipt: sliced gate census (49 files) ⊇ matrix files (18),
+  31 files of extra coverage. `needs: evals` edge dropped, PR comment moved
+  into slices-report, KNOWN_MATRIX_GAPS/KNOWN_TIER_UNSET retired,
+  test/evals-workflow-matrix.test.ts rewritten as
+  test/evals-workflow-wiring.test.ts. The register-skills fail-fast
+  verification loop was ported to the surviving lanes FIRST via the shared
+  .github/actions/register-gstack-skills composite.
 - **P1 — Maintainer decision: make `slices-report` a required check** once
   post-migration flake data exists (the Codex outside-voice's "green means
   green is not delivered while paid stays advisory" point — correct, and