OPEN SOURCE · MADE FOR YOUR EXISTING AGENTS

Your agents
A new home

Bring the agents and skills you already wrote to Cursor Agent SDK. Keep the source. See every tradeoff. Build from there.

Read the code. Make it yours.

Install the CLIv0.1.1
npm install --global @kyun.si/agent2cursor@0.1.1
  • MIT licensed
  • Node.js ≥22.13
  • No login to convert
BRING THE WORK YOU ALREADY DID
Claude Code+Codex+Agent Skills→Cursor Agent SDK

01 / A BRIDGE, WITH THE DETAILS

A head start you can inspect

Go from scattered agent files to a project you own. The migration report stays beside the code, so the next decision is yours.

01

Source you can keep building on

Get agent.ts, original instructions, skills, and supporting files in a standalone TypeScript project. Open it, change it, commit it.

02

Nothing quietly lost in translation

Every source field is recorded as mapped, degraded, or dropped. Model choices, tool restrictions, and unsupported hooks are explicit.

03

A starting point for verification

Generated routing evals help check which agent or skill is selected. Install the target dependencies to run structural validation; test real behavior separately.

INSIDE THE GENERATED PROJECT
my-agent/
├─ bot/
│  ├─ agent.ts
│  ├─ instructions.md
│  ├─ subagents/
│  │  └─ code-reviewer/
│  │     ├─ agent.ts
│  │     └─ instructions.md
│  ├─ skills/
│  └─ tools/
├─ evals/
├─ MIGRATION.en.md
└─ package.json
Source instructions are preserved, not translated.
See a real fixture conversion
Claude Code, Codex and Agent Skills become an editable Cursor Agent SDK project and a migration report.

Repository fixture → generated source and field report. Structural validation is separate from a live model run.

02 / FROM YOUR TERMINAL

Start with one agent

Install the CLI above, then run these commands in a project with the selected input. Use a path that exists and a new output directory.

01

Inspect before writing

Read the migration plan. This command writes nothing.

terminal
agent2cursor inspect .claude/agents --lang en
02

Generate an editable project

Creates ./my-agent and a migration report. Existing output is never overwritten.

terminal
agent2cursor convert .claude/agents --out ./my-agent --lang en
03

Review, then validate

First review the report, model choices, and permissions. These commands validate structure and list evals; they do not run a model.

my-agent/
cd my-agent
npm install
npm run validate
npx agent-sdk eval --dir . --list

03 / GOOD TO KNOW

A few things before you move

What can I bring?

Claude Code Markdown subagents, standalone Codex TOML agents, SKILL.md directories and their supporting files, and tstack skill workflows. Inputs can be combined. Parent config.toml, credentials, and chat history are not imported.

Why not just copy SKILL.md?

If you only need compatible skills, copying them may be enough. agent2cursor is useful when you also want subagent conversion, editable SDK scaffolding, field-by-field reports, disabled script adapters, and routing evals.

Will it behave exactly the same?

No. The report makes differences visible, but it cannot guarantee equivalent model behavior or runtime permissions. Source model IDs inherit the target parent by default. Review the generated project and test your own tasks.

Do I need a Cursor account?

Not to inspect or convert. Post-install structural validation also needs no login. Live model runs and actual model evals require a Cursor account with model access. Hosting an agent has separate account and repository requirements.

Is this an official Cursor project?

agent2cursor is an independent, MIT-licensed project by Eason Yang. It is not affiliated with or endorsed by Cursor, Anysphere, Anthropic, OpenAI, xAI, Theo Browne, or T3. The target SDK is a separate dependency.

SMALL TOOL. OPEN SOURCE. YOURS TO BUILD ON.

Keep good agents moving

Useful today, or for your next migration? Give it a star so it’s easy to find again. Ideas, issues, and pull requests are welcome.