test(e2e): Tier C — meld→kiln execution seam (#297), #[ignore]d on kiln#364

[avrabe]

What

Closes the test side of #297 — adds Tier C to golden_e2e.rs: execute a meld-fused component on kiln (the MCU-target runtime), not just wasmtime (Tier A/B). This is the "honest boundary" Tier A/B explicitly disclose ("NOT on the synth/kiln MCU target").

The seam is currently broken — kiln#364

Standing up the test immediately surfaced a real meld→kiln interop blocker (grounded repro on #297):

• meld's fused --component output preserves the exact component WIT (export wasi:cli/run@0.2.6) and runs under wasmtime via the canonical entry (Tier A/B green).
• kiln can't execute it: its component runner looks for a core-instance _start, which meld's multi-core-module wrap (stubs / fused / fixup / caller, deferred start) doesn't expose → ✗ No core instance exports _start.

Filed kiln#364 (suggested fix: run components via their wasi:cli/run canonical entry, like wasmtime).

This PR

• Tier C fuses a wasi:cli command fixture, runs it on kilnd, and asserts execution — with the unfused original as the on-kiln baseline so a fused failure is attributable to fusion (not a missing-fixture/kiln-absent case).
• #[ignore]d with the kiln#364 reason, so normal CI stays green. Verified it fails for exactly the documented cause (baseline ran on kiln, fused did not). Un-ignore when kiln#364 lands and it pins the meld→kiln seam green.
• kilnd located via MELD_KILND or the conventional ../../kiln debug build; absent → skip.

Normal golden_e2e suite: 4 pass, Tier C ignored. clippy clean.

Refs: #297, kiln#364.

🤖 Generated with Claude Code

[github-actions]

LS-N verification gate

✅ 58/58 approved LS entries verified

	count
Passed (≥1 test, all green)	58
Failed (≥1 test failure)	0
Missing (no ls_*_NN_* test found)	0

_{Approved loss-scenarios.yaml entries are expected to have a
regression test named ls_<letter>_<num>_* (e.g. LS-A-11 →
ls_a_11_*). The gate runs each prefix via cargo test --lib --no-fail-fast and aggregates pass/fail/missing.}

Failed LS entries

(none)

Missing regression tests

(none)

_{Updated automatically by tools/post_verification_comment.py.
Source of truth: safety/stpa/loss-scenarios.yaml.}