Skyvern for AI agents
Skyvern runs browser-based automation with AI agents that navigate real websites, fill forms, and extract data for tasks that don't have a clean API. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Skyvern.
Start a browser automation task
Start a browser task that navigates a target site and completes a defined action, like a form submission or checkout.
Check a run before trusting its output
Get a run's status and results before treating the extracted data as final, catching failed navigations early.
Cancel a stuck run
Cancel an in-progress run that is taking too long or heading toward the wrong page, before it takes any action.
10 Skyvern actions, graded by risk.
Every Skyvern action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
5Look things up. Allowed by default.
- skyvern.get_runGet Run
- skyvern.list_runsList Runs
- skyvern.get_folderGet Folder
- skyvern.list_agentsList Agents
- skyvern.list_foldersList Folders
write
3Create and change things. Allow, or ask the user first.
- skyvern.create_folderCreate Folder
- skyvern.update_folderUpdate Folder
- skyvern.start_browser_taskStart Browser Task
destructive
2Delete, cancel or archive. Ask first, or deny outright.
- skyvern.cancel_runsCancel Runs
- skyvern.delete_folderDelete Folder
Skyvern in three steps.
- 01Your users connect SkyvernThey add their Skyvern api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “create Folder” can wait for the user, and “delete Folder” can be denied outright.
- 03Any agent can actYour agent calls Skyvern through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('skyvern', { read: 'allow', write: 'ask', // create_folder destructive: 'deny', // delete_folder }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Skyvern connects.
Users add their Skyvern api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Skyvern connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Skyvern from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Skyvern and Arc0, answered.
Can I use Skyvern with Claude, ChatGPT or Cursor?
Yes. Connect Skyvern to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Skyvern actions you allow.
How do users connect Skyvern?
Users add their Skyvern api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Skyvern actions can my agent take?
10 in total: 5 read, 3 write and 2 destructive, such as “create Folder”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in Skyvern?
Yes. Actions like “delete Folder” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call Skyvern too?
Yes. The same Skyvern connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Skyvern into your agent.
Your users connect Skyvern once, under your brand. Your agent gets 10 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan