CodePlane¶
Run coding agents. Supervise from anywhere.
No IDE. No terminal. Just a prompt.
CodePlane runs Claude Code and GitHub Copilot on your workstation — supervise from any browser on your desktop, phone, or tablet. Review diffs, approve risky actions, track costs, and merge when you're ready.
Works with Claude Code CLI and GitHub Copilot CLI · Open source, MIT license
The Core Loop¶
1
Launch a task¶
Pick a repository, write a prompt, choose an agent and model. The agent runs in an isolated Git worktree — your working directory is never touched.
2
Supervise the run¶
Watch the transcript, logs, plan progress, and cost data while the agent works. Send messages to steer it if needed.
3
Gate risky actions¶
File writes, shell commands, and destructive operations can require your approval before they execute.
4
Land or discard¶
Review the diff, then merge, create a PR, or discard — based on what the agent actually produced.
Supported Agents¶
CodePlane works with GitHub Copilot CLI and Claude Code CLI. Install and authenticate either CLI, select your agent and model per job — CodePlane manages the underlying SDKs and handles the rest.
External agents can orchestrate CodePlane programmatically through its built-in MCP server — compatible with VS Code, Claude Desktop, Cursor, and any MCP-compatible client.
What You Get¶
Task Orchestration¶
Launch jobs with a prompt and model selection. Each job runs in its own Git worktree for safe, concurrent execution.
Mobile-First & Remote¶
Run on your workstation, control from any browser — phone, tablet, or desktop. UI is touch-optimised. Remote access out of the box via Dev Tunnels or Cloudflare Tunnels.
Live Visibility¶
Transcript, logs, timeline, plan steps, and token costs — all streaming in real time as the agent works.
Approval Gates¶
Risky operations pause for your review. Approve, reject, or trust the session to auto-approve the rest.
Diff Review & Merge¶
Syntax-highlighted diffs, workspace browsing, and merge/PR/discard controls — all built in.
Cost Analytics¶
Track token usage, costs, model performance, and tool health across all jobs.