Complete SOW-0050 and SOW-0058 backend gate hardening

[ktsaou]

Summary

• Completes SOW-0050 first backend residual-complexity slice and closes the SOW ledger.
• Completes SOW-0058 benchmark gate stabilization, including deterministic benchmark behavior, retry diagnostics, and CI parser hardening.
• Addresses PR check findings in ingest tests and benchmark shell checks without weakening assertions.
• Adds SOW-0059 to track the remaining pre-existing ingest test complexity inventory separately from this PR.

Validation

• timeout 3600 ./scripts/gates.sh passed after PR check remediation in 1581s.
• Benchmark regression gate self-test passed 41/41; real benchmark attempt 1 reported Tail_SyntheticAppend measurement noise at +20.06% sec/op; attempt 2 did not reproduce that same benchmark-name regression, so the benchmark gate passed by its retry contract.
• Focused ingest tests passed with and without -race.
• bash -n scripts/check-bench.sh scripts/test/check-bench-test.sh, ShellCheck, and the benchmark self-test passed.
• Local Codacy analysis on touched files reported zero ShellCheck, Semgrep, Trivy, and benchmark shell-script issues; remaining Lizard output is pre-existing ingest test/function/file-size complexity tracked by SOW-0059.
• Go race/coverage passed with total statement coverage 86.0% and gated internal/* aggregate 91.0%.
• Frontend Vitest passed 631/631 with 94.41% statement coverage.
• Playwright E2E and axe passed 51/51.
• Spec drift, secret scan, repository attribution scan, Codacy self-tests, systemd lint, build, bundle gate, and adapter fuzz seed corpus passed.
• Final independent review convergence was codex + kimi + deepseek, all PRODUCTION GRADE: YES on the final staged diff.

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

🟢 Metrics 140 complexity · 26 duplication

Metric	Results
Complexity	140
Duplication	26

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}

[cubic-dev-ai]

1 issue found across 23 files

<sub>Tip: cubic can generate docs of your entire codebase and keep them up to date. Try it here.

Re-trigger cubic</sub>

[ktsaou]

Addressed the benchmark failure-diagnostic finding by capturing the failed benchmark command status inside the failing branch and tightening the benchmark self-test to require the fake go test exit code. Added SOW-0059 to track the remaining pre-existing ingest test complexity inventory and documented the Codacy/Lizard count split.