chore(release): prepare v0.15.4

[GabrielePicco]

Summary by CodeRabbit

• Chores
  • Updated the package release version to 0.15.4.
  • Updated platform-specific optional package versions to 0.15.4.
  • Synchronized the workspace and package template versions with the new release.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-18T12:11:21.812819Z	3564630	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Advanced

Run ID: 60559ee7-ef66-4b13-b6be-7f80da1c45f5

📥 Commits

Reviewing files that changed from the base of the PR and between 2fd6477 and 3564630.

⛔ Files ignored due to path filters (2)

• Cargo.lock is excluded by !**/*.lock
• test-integration/Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (3)

• .github/packages/npm-package/package.json
• .github/packages/npm-package/package.json.tmpl
• Cargo.toml

Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

The workspace and npm package versions are updated from 0.15.3 to 0.15.4. The four platform-specific optional validator dependencies and the npm package template are also updated to 0.15.4.

Suggested reviewers: bmuddha

Priority: ⬇️ Low

Merge Risk: ⚪ Minimal · up to 35646

The release metadata is aligned for the 0.15.4 publication workflow, with no identified merge-blocking risk.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

redsuite: PR vs master

Single-run diff on shared runners — indicative only; statistical verdicts come from Bencher thresholds.

redline/clone_lru_churn/cap360
  read latency us                    median 206 → 611 (+196.6%)  p95 1891 → 4337 (+129.3%)  ▲ worse
  (8 flat/mixed/info metric(s) not shown)

redline/clone_lru_churn/cap50
  read latency us                    median 1693 → 7525 (+344.5%)  p95 7911 → 19048 (+140.8%)  ▲ worse
  (8 flat/mixed/info metric(s) not shown)

redline/clone_lru_churn/closure
  read latency us                    median 178 → 507 (+184.8%)  p95 289 → 1040 (+259.9%)  ▲ worse
  (8 flat/mixed/info metric(s) not shown)

redline/ensure_gate_stall/healthy
  delivery us                        median 250 → 511 (+104.4%)  p95 423 → 963 (+127.7%)  ▲ worse
  (6 flat/mixed/info metric(s) not shown)

redline/rpc_warm_ingress
  delivery us                        median 444 → 166 (-62.6%)  p95 722 → 303 (-58.0%)  ▼ better
  validator tx processing avg us     155.9 → 71.4 (-54.2%)  ▼ better
  validator ensure accounts avg us   56.7 → 22.4 (-60.4%)  ▼ better
  (9 flat/mixed/info metric(s) not shown)

redline/storage_prodsize_sustain/frequent_snapshot
  window A delivery us               median 155 → 648 (+318.1%)  p95 348 → 1126 (+223.6%)  ▲ worse
  window B delivery us               median 126 → 547 (+334.1%)  p95 217 → 915 (+321.7%)  ▲ worse
  window A validator tx avg us       70.6 → 207.4 (+193.8%)  ▲ worse
  window B validator tx avg us       59.5 → 176.1 (+196.2%)  ▲ worse
  (5 flat/mixed/info metric(s) not shown)

redline/storage_prodsize_sustain/prod_cadence
  window B delivery us               median 211 → 496 (+135.1%)  p95 415 → 1021 (+146.0%)  ▲ worse
  (8 flat/mixed/info metric(s) not shown)

9 metric(s) worse than base

[github-actions]

Bencher Report

Project	magicblock-labs
Branch	release/v0.15.4
Testbed	blacksmith-8vcpu-ubuntu-2404

⚠️ WARNING: Truncated view!

The full continuous benchmarking report exceeds the maximum length allowed on this platform.

🚨 17 Alerts

🐰 View full continuous benchmarking report in Bencher