OmniTools

A local-first Chrome developer toolbox that keeps everyday workflows fast, unified, and private.

Comment

Inspiration

Developers often switch between many websites, desktop apps, and browser extensions for small but frequent tasks: inspecting headers, formatting JSON, testing APIs, converting encodings, processing images, reading QR codes, and checking colors. This context switching is slow, fragmented, and can expose sensitive data to remote services.

What it does

OmniTools is a local-first Chrome Manifest V3 developer toolbox with a unified, searchable workbench.

Current tools include:

  • Request and response Header debugging with reusable profiles and URL/method/resource filters
  • JSON formatting, validation, sorting, statistics, and structural diff
  • A lightweight REST API client with auth, environments, history, collections, and cURL import/export
  • 24 encoding, decoding, hashing, and structured-inspection operations
  • A local image workbench for batch conversion, crop/rotate/flip, high-quality resize, target-size compression, watermarks, EXIF inspection/removal, Base64 conversion, and ZIP export
  • QR code generation/recognition and a screen color picker

Data stays in the browser by default. There are no accounts, telemetry, ads, or remote synchronization.

How it works

OmniTools uses Vue 3, TypeScript, Vite, and Chrome Manifest V3. A central tool registry powers the popup launcher and full workbench. Larger modules are lazy-loaded, and each feature keeps its UI, data, permissions, and implementation inside its own feature boundary.

Browser-native APIs and local libraries do the work: Declarative Net Request for header rules, Fetch for API testing, Canvas and Pica for image processing, JSZip for batch export, and Exifr for local metadata inspection.

How Codex and GPT-5.6 were used

OmniTools was built and refined with Codex powered by GPT-5.6. Codex helped trace the real extension architecture, turn product requirements into scoped feature modules, implement the local image-processing pipeline, harden Chrome DNR edge cases, add type-safe workbenches, and verify the project through type checking, production builds, and browser-based interaction checks.

Key engineering decisions made with Codex included keeping sensitive workflows local-first, using a composable image pipeline instead of fragmented tools, lazy-loading heavy features, and enforcing Chrome's request-header append allowlist across storage, UI, and compilation.

Challenges

The most demanding parts were browser-platform constraints: safe Header mutation under Manifest V3, target-size image compression without server processing, duplicate-safe batch ZIP export, and maintaining a compact experience while supporting full-featured tools.

What's next

Before final submission, we will add a judge-friendly installation package, English setup and testing instructions, sample workflows, final browser verification, and a concise public demo showing how Codex and GPT-5.6 accelerated the build.

Built With

Share this project:

Updates