Blog
Notes on Claude Code, the Agent SDK, and shipping AI software.
- Jul 2, 2026 · 8 min read
Claude Code GUI: CLI vs official Desktop vs IDE plugins vs Cockpit (2026 buyer’s guide)
Anthropic ships Claude Code as a CLI by default. If you want a GUI on top, you have five real options today: stick with the CLI, use the official Claude Code Desktop app, an IDE plugin (Cursor, Continue), Aider in a TUI, or run Cockpit. Here is when each one wins.
Read more → - Jul 2, 2026 · 2 min read
Cockpit now takes an access token
Cockpit is local-first and fully open on your own machine. When you expose it on a LAN or a cloud sandbox, an optional access token now guards remote access — while local stays frictionless.
Read more → - Jul 2, 2026 · 6 min read
Self-Host a Claude Code GUI for Your Whole Team (One Dev Box, Zero Installs)
OpenCockpit is web client–server: install it once on a shared dev box, and every teammate opens a browser and gets a seat — each coding with AI in their own project or worktree. Setup takes 5 minutes.
Read more → - Jun 23, 2026 · 2 min read
Cockpit on your phone
Cockpit now has a mobile layout. Open it on your phone to see your sessions, open any chat, and receive a notification when a run finishes.
Read more → - Jun 18, 2026 · 4 min read
Orchestrate a workflow from the message box
A Cockpit message can now hold more than one command. Start several lines with `/` or `@` and Cockpit reads the whole message as an ordered, multi-step workflow — `/` runs in the main session, `@` delegates to a sub-agent. Plan the run once, send it once.
Read more → - Jun 10, 2026 · 5 min read
Cockpit now has a Claude Code CLI execution mode
Every Claude chat tab in Cockpit can now run in one of two ways: the Claude Agent SDK (headless `query()`), or a real interactive `claude` session driven in a PTY. Both share the same conversation, so you can switch per tab, mid-conversation, one click away. Here is what the new CLI mode is and how it works.
Read more → - May 22, 2026 · 6 min read
What is a Code Graph (and why your AI needs one)
A code graph is the structured map of your project's symbols and their relationships — who calls whom, what depends on what, which files always get edited together. It is exactly the missing layer between an AI agent and your codebase, and the reason `grep` is the agent's ceiling on the questions that matter most.
Read more → - May 12, 2026 · 6 min read
Vibe coding needs a bit of taste
Vibe coding's hidden cost isn't speed — it's entropy. Cockpit's repo today is two piles: `packages/feature/` is the business, `packages/shared/` is the common floor, arrows go one way. From that picture, three old-school engineering habits — putting things in the right place, drawing real boundaries, deleting what doesn't fit — become more valuable when the agent is the one typing.
Read more → - May 7, 2026 · 7 min read
Read code as a map, not a tree
The file tree shows you where bytes are stored. It does not show you how the code actually moves. Cockpit's new Code Map turns any source file into a canvas of function chips — callers on the left, callees on the right, click a pin to jump. Five clicks across an unfamiliar repo and you've walked the auth flow. Here is what that looks like in five real scenarios.
Read more → - Apr 30, 2026 · 6 min read
From chat to skill: turn yesterday’s 28-minute debug into today’s /command
Every productive Claude Code session ends with dark knowledge that dies with the conversation. Cockpit’s Skills feature lets the agent crystallize a chat into a SKILL.md — saved to *your* knowledge base (not a Cockpit-owned folder), then registered as a slash command. Your notes stay where they live; Cockpit just holds the pointer.
Read more → - Apr 30, 2026 · 4 min read
Use DeepSeek inside Cockpit — and keep all your Claude habits
Cockpit now talks to DeepSeek. Open a tab, paste a key, and DeepSeek-v4 edits your files, runs your terminal, reviews your diffs — exactly the way you already use Claude. Here is how to set it up in under a minute and what to expect.
Read more → - Apr 29, 2026 · 6 min read
How to run 5 Claude Code sessions in parallel without losing your mind
Claude Code is incredible at one task at a time — but most engineers want to scope out three features while one is refactoring and another is writing tests. Here is how Cockpit lets you run multiple Claude Code Agent SDK sessions across projects at once, without context-switching pain.
Read more →