Skillget

cockpit

agent
Guvenlik Denetimi
Basarisiz
Health Uyari
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 6 GitHub stars
Code Basarisiz
  • process.env — Environment variable access in bin/cock-browser.mjs
  • network request — Outbound network request in bin/cock-browser.mjs
  • process.env — Environment variable access in bin/cock-dev.mjs
  • process.env — Environment variable access in bin/cock-terminal.mjs
  • network request — Outbound network request in bin/cock-terminal.mjs
  • exec() — Shell command execution in bin/cock.mjs
  • spawnSync — Synchronous process spawning in bin/cock.mjs
  • process.env — Environment variable access in bin/cock.mjs
  • network request — Outbound network request in bin/cock.mjs
  • execSync — Synchronous shell command execution in bin/postinstall.mjs
  • process.env — Environment variable access in bin/postinstall.mjs
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This is a graphical user interface designed to run parallel AI coding sessions using the official Claude Agent SDK. It provides a local-first dashboard with built-in terminal access, browser control, and database management.

Security Assessment
Risk Level: Medium to High. The tool is designed to orchestrate complex development tasks, which inherently requires high system permissions. The automated scanner flagged multiple instances of shell command execution (`exec()`, `spawnSync`, `execSync`), which is expected given its deep integration with your terminal and the Claude Code environment. The code makes outbound network requests to communicate with APIs and relies on environment variables to manage credentials safely. No hardcoded secrets or dangerous requested permissions were found. However, the most significant vulnerability is the `postinstall.mjs` script, which automatically executes shell commands during the `npm install` process. Running unverified installation scripts from a low-visibility project always carries the risk of executing malicious code on your machine.

Quality Assessment
The project is very new and currently has minimal community adoption with only 6 GitHub stars. Because of this low visibility, it has not been widely audited by independent security researchers, meaning you are placing a high degree of trust solely in the original developers. On a positive note, it is actively maintained (last updated today), has a clear description, and uses the standard, permissive MIT license.

Verdict
Use with caution. While its architecture and features are exactly what developers need for complex AI workflows, its minimal community presence combined with highly privileged automatic post-install shell scripts means you should thoroughly review the installation code before running it on your machine.
SUMMARY

Claude Code GUI for parallel AI coding — multi-project sessions, terminal, browser & DB bubbles, code review, slash modes. Built on the official Claude Agent SDK. Local-first, MIT.

README.md

Cockpit logo

Cockpit — A Claude Code GUI for parallel AI coding

One seat. One AI. Everything under control.
/ˈkɒkpɪt/ — like an aircraft cockpit

npm version npm downloads GitHub stars MIT license website Built on Claude Agent SDK

English · 中文 · Website · Blog

Cockpit is an open-source Claude Code GUI. Run multiple Claude Code Agent SDK sessions in parallel across projects, with a built-in terminal, Chrome control, PostgreSQL / MySQL / Redis bubbles, code review, and slash modes — all local, zero config.

https://github.com/user-attachments/assets/18f1a5dc-64f3-4ff6-b9fc-9cd08181fbb8

npm i -g @surething/cockpit && cockpit

Why Cockpit?

Anthropic ships Claude Code as a CLI. That's the right call for power users — but the moment you have more than one project in flight, your terminal turns into mission control with no instruments.

Cockpit is the instrument panel. It does not replace Claude Code; it stands on top of the official Agent SDK and gives you the things a CLI can't:

Pain with raw Claude Code What Cockpit adds
One session at a time, terminal chaos at 3+ projects Multi-project tabs, parallel agent sessions, red-dot inbox, desktop notifications
Image attachments are awkward Drop / paste images straight into chat
"What was I debugging yesterday?" Cmd+K cross-project session browser, pinning, forking
Agent can't reach your browser / DB Smart Bubbles: Chrome, PostgreSQL, MySQL, Redis — drivable by the agent
Reviewing AI output is friction LAN-shared review pages, line-level comments, send any comment back as AI context
Same "do X but don't change code" prompt every day Slash modes /qa /fx /review /commit + custom ~/.claude/commands/*.md
No automation hooks One-time / interval / cron-based scheduled tasks
"Cloud relay" trust concerns Fully local. No telemetry. No API key beyond what claude already has.

Features

Agent — AI chat that scales

  • Powered by the official Claude Agent SDK — zero extra API key setup
  • Multi-project concurrent sessions with desktop notifications and red-dot badges
  • Session pinning, forking, cross-project session browser (Cmd+K)
  • !command prefix to run shell from chat — output piped back as context
  • Image attachments, code references, token usage tracking

Explorer — Code & files

  • 4-tab file browser: Directory · Recent · Git Changes · Git History
  • Syntax highlighting (Shiki) with Vi mode editing
  • Git blame, diff view, branch switching, worktree management
  • LSP integration — go to definition, find references, hover info
  • Fuzzy file search (Cmd+F), JSON viewer, Markdown preview

Console — Terminal & smart Bubbles

  • Full xterm.js terminal with shell integration
  • 🌐 Browser Bubble — control Chrome via accessibility tree (click, type, navigate, screenshot, network)
  • 🐘 PostgreSQL Bubble — browse schema, run queries, export
  • 🐬 MySQL Bubble — browse databases & tables, run queries
  • 🔴 Redis Bubble — browse keys, inspect values, execute commands
  • Drag-to-reorder, grid / maximized layout, per-tab env vars & shell aliases

Code Review — LAN-shared, no SaaS

  • LAN-shareable review pages — teammates need zero install
  • Line-level comments with reply threads
  • Send any comment back to AI as context for an automated fix
  • Red-dot badges keep unread feedback visible across projects

Slash modes — change the agent's posture

  • /qa — Clarify-only: restate, ask back, never code
  • /fx — Diagnose-only: bug evidence chain, never edit
  • /review — read the diff, write notes, no rewrites
  • /commit — stage + draft a message in your repo's style + commit
  • Custom: drop any *.md into ~/.claude/commands/ or ./.claude/commands/ → instant slash command

Scheduled tasks — Cron for AI

  • One-time, interval, or cron scheduling
  • Pause / resume, reorder, track results across projects

Skills — extensibility

  • Drop in any SKILL.md to teach the agent a new trick
  • Invoke with /skill-name from chat
  • All managed from a single Skills sidebar

Use cases

  • Solo dev, multi-repo: "I have a refactor running in API, tests writing in Web, and a bug investigation in Pipeline — all at once, all visible."
  • Two-person team: Senior reviews via LAN-shared review page, no GitHub PR round-trip needed for in-progress work.
  • Full-stack chore mode: /fx in one tab on a backend bug, /review in another on the frontend diff, /commit to wrap up — three slash modes, three different agent postures.
  • AI-driven QA: Browser Bubble + scheduled task = "every night at 2 AM, run this UI smoke flow and post a summary".
  • Privacy-sensitive code: runs on your laptop, talks only to the Claude API your claude CLI is already configured with. No telemetry, no relay.

Try online

No install, no AI chat (read-only sandbox, 5 min):

Try Online

Prerequisites

  • Node.js ≥ 20 — nodejs.org
  • Claude Code — docs.anthropic.com/en/docs/claude-code
  • Git — for git features (blame, diff, worktree, etc.)
  • Chrome (optional) — for Browser Bubble; install the bundled extension from ~/.cockpit/chrome-extension

Install

npm install -g @surething/cockpit
cockpit                # start cockpit → http://localhost:3457
cockpit .              # open current dir as a project
cockpit ~/my-project   # open specified dir
cockpit -h             # help

Both cockpit (full name) and cock (short alias) ship with the package — use whichever you prefer. Docs and examples use cockpit; existing muscle memory keeps working.

From source

git clone https://github.com/Surething-io/cockpit.git
cd cockpit
npm install
npm run setup       # build + npm link (registers `cockpit` and `cock`)

CLI

cockpit browser <id> snapshot      # capture accessibility tree
cockpit browser <id> click <uid>   # click element
cockpit terminal <id> exec "ls"    # execute command
cockpit terminal <id> output       # get terminal output

Comparison

Raw Claude Code CLI IDE plugin (Cursor, Continue) Aider TUI Cockpit
Multi-project parallel tmux required multi-window one at a time first-class
Cross-project search grep per-window local Cmd+K
Browser / DB control ❌ usually ❌ ❌ ✅ Bubbles
Code review surface git tools PR provider git LAN-shared
Slash modes manual per-plugin yes /qa /fx /review /commit + custom
Local-only / no cloud relay ✅ varies ✅ ✅
Day-1 SDK features ✅ wait varies ✅ official SDK
Open source ✅ mostly ❌ ✅ ✅ MIT

Read the long version: Claude Code GUI: CLI vs Cockpit vs IDE plugins

Read more

Development

npm run dev         # dev server → http://localhost:3456
npm run build       # production build
npm run setup       # build + npm link
npm run lint        # ESLint

Tech stack

Next.js 16 · React 19 · TypeScript · TailwindCSS · xterm.js · Shiki · i18next · Claude Agent SDK

Contributing

Issues and PRs welcome. See CONTRIBUTING.md and GUIDE.md.

License

MIT © Surething

If Cockpit saved you 10 minutes today, a ⭐️ on GitHub is the cheapest thank-you we know.

Yorumlar (0)

Sonuc bulunamadi