neuron-explorer
Steering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout, browse the results
Bridge is a native desktop control room for supervised coding-agent work. It runs Codex, Claude Code, and OpenCode as one team, with a policy engine on every gate and a Git worktree for every worker.
Developer toolsRust · TypeScriptGitHub @cestercianOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
The last coding agent you’ll ever need.
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
The last coding agent you'll ever need.
Claude Code, Codex, Cursor, OpenCode, and Grok, running as one team in one window, on your own subscriptions.
Ship more, babysit less.
Bridge is a native desktop app for running AI coding agents on real projects. Point it at a Git repository, start the agents you want, and let them work in parallel while you review what lands. Your code and your credentials stay on your machine.Read the full README ↗Bridge is a native desktop app for running AI coding agents on real projects. Point it at a Git repository, start the agents you want, and let them work in parallel while you review what lands. Your code and your credentials stay on your machine.
Run as many agents as you like on one repo. Every chat gets its own Git worktree and branch, so they never step on each other's files. Mission Control shows them side by side and flags the one waiting on you.
Change model or provider inside one conversation, Codex to Claude Code to Cursor, without starting over. The provider session restarts; your history stays where it is. Nobody else lets you do this.
When an agent needs a site you're logged into, it asks. Approve it and the agent drives a throwaway copy of your session: it can click through a deploy preview or a dashboard, and you can take over at any point. The copy is thrown away when the turn ends.
Every message, plan, tool call, and delegation lands in a local log that only ever grows. When the context window fills, Bridge compacts the model's view and leaves the log alone, so a restart picks up from a verified checkpoint instead of a blank page.
Tests run first. Then a reviewer from a different model family has to sign off, and a review from the family that wrote the code is thrown out. Claude's work gets checked by Codex, and the other way round.
Preferences, decisions, and constraints get saved as you work and come back in any later chat, whichever agent is running and whichever repo it's in. Every reply shows which memories it used.
Narrow work goes to a cheap tier and only the hard parts reach an expensive one. Tokens, cost, and cache savings break out per harness and per model, so the routing pays for itself visibly.
| Agent | CLI | Bridge can install it |
|---|---|---|
| Codex | codex | Yes |
| Claude Code | claude (needs Node.js 18+) | Yes |
| OpenCode | opencode | Yes |
| Cursor | cursor-agent | Yes |
| Grok | grok | Install it yourself |
Sign-in always runs the vendor's own login command. Bridge never collects, stores, or logs your provider credentials.
Missing an agent? It simply shows as unavailable — Bridge still starts and everything else keeps working. Your own installs always take precedence over the ones Bridge manages.
.dmg, drag Bridge into Applications, and launch it.Approvals, history, and usage tracking are on from the start.
Bridge_*.dmg, signed with a Developer ID, notarized, and stapled[!NOTE] Bridge is early-stage software. Expect rough edges and frequent improvements. macOS may ask for file access the first time Bridge touches
Desktop,Documents, orDownloads— keeping repos in a folder like~/Codeavoids repeated prompts.
Bridge is a Tauri 2 app with a Rust workspace under src-tauri/ and a React + TypeScript + Vite frontend styled with Tailwind CSS v4, managed with Bun.
See CONTRIBUTING.md for prerequisites, the full command table (bun run dev, bun run tauri dev, bun run check, bun run build, bun run test), pull request expectations, and issue etiquette (docs/issue-format.md — every issue needs For humans and For agents sections).
Agent and UI conventions: AGENTS.md.
git clone https://github.com/Atharva-Kanherkar/bridge-harness.git
cd bridge-harness
bun install
bun run dev # fast frontend iteration (mock data, no Rust shell)
bun run tauri dev # full desktop app
Before opening a PR:
bun run build
bun run test
Keep changes focused and follow AGENTS.md (Tailwind CSS v4 only,
colocated tests). PR titles must use Conventional Commit syntax (feat:,
fix:, docs:, chore:), and PRs are squash-merged so that title becomes the
release commit. fix: releases a patch, feat: a minor, and ! marks a major
version.
Release Please maintains one release PR. Squash-merging that PR creates a version tag and draft release; production CI signs and notarizes the exact tagged build, runs the packaged app and its bundled daemon from the mounted DMG in a credential-free job, and publishes only after every gate passes. See the macOS release guide for credentials, verification, and recovery.
Bridge is open source under the MIT License. Use it, fork it, ship it.
Third-party code carries its own terms; see THIRD_PARTY_NOTICES.md.