Hosted MCP connector

Percher runs a hosted, OAuth-protected MCP server at https://mcp.percher.app/mcp. You add it to Claude.ai (and Cursor, Windsurf, and other clients) as a custom connector— there's nothing to install and no token to paste. The client runs the OAuth login for you, and you approve a Percher consent screen once.

It lets an AI assistant operate your Percher apps and publish an app it just built for you— say “publish my app” in Claude and the app it generated goes live. The one thing the hosted server can't do is read files on your local disk; for an existing project on your machine, use the local package (see below).

How to add it

• Claude.ai (web/desktop):Settings → Connectors → Add custom connector → https://mcp.percher.app/mcp→ approve the Percher login. (Custom connectors require a paid Claude plan.)
• Cursor / Windsurf: add it as a remote MCP server by URL; they run the OAuth flow for you.
• ChatGPT / Codex, or any stdio-only client: bridge it with npx -y mcp-remote https://mcp.percher.app/mcp — mcp-remote handles the OAuth dance and presents a local stdio server the client can talk to.

npx -y mcp-remote https://mcp.percher.app/mcp

What the hosted connector can do

The hosted server runs against the platform, not your local disk — it's the remote-safe subset of the tool set:

• percher_publish — publish an app from source the assistant passes inline: the files it generated in your session (include a percher.toml). Percher packs them server-side, creates the app, and deploys. This is the hosted “I just said publish” path for an assistant with no local file access. Inline source is capped at ~25 MB; for a larger app, connect a repo.
• percher_deploy — deploy the latest commit of a repo you've connected via the GitHub App. Percher clones it server-side. percher_redeploy re-runs the current version.
• List apps and app status
• Env get / set / list
• Custom domains
• Deploys: list + inspect
• Version history and diffs
• Public inspect links
• Backups and data status
• Resume (unsuspend) an app

Manage or revoke connected clients in the Percher dashboard under Account → Connected apps.

Hosted connector vs the local package

The hosted connector can publish source the assistant hands it inline (percher_publish) or a connected repo (percher_deploy). What it can't do is read files on your local disk— a project folder the assistant can't see, a large app, or anything you'd publish from a terminal. For that, use the local MCP server npx -y @percher/mcp (or the CLI bunx percher publish), which reads your project directory.

• Local package (@percher/mcp / bunx percher publish) — publish from files on your machine, full tool set, runs on your machine.
• Hosted connector (mcp.percher.app/mcp) — publish an app the assistant built (inline) or a connected repo, operate existing apps, zero install.

For the local setup, see the MCP tools and Deploy instruction for AI agents sections.

Security

The hosted connector uses OAuth 2.1 with a per-user login. Tokens are scope-limited and revocable in real timefrom the dashboard — revoking a client disconnects it immediately. Publishing is bounded by your plan's app and deploy limits, the same as the CLI.

Because any client can self-register (open Dynamic Client Registration), the consent screen shows the exact redirect origin the approval is sent to, plus an "unverified client" label. Check that origin before you approve.