Publish static websites to a live HTTPS URL from your AI tool. Free tier, instant deploys.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Bearer p5_... token from https://my.press5.xyz/launch. Optional: the tools can sign you in by email.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Delete a site and all its deploys. It goes offline at once and the name becomes free for anyone. Ask the person first.
Must be true
A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check.
The server marks this tool as destructive. Running it may delete or overwrite data on the server or in any account you connected with your headers.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http launch-pad https://my.press5.xyz/mcp
Headers
Authorization (optional, not in the config above): Bearer p5_... token from https://my.press5.xyz/launch. Optional: the tools can sign you in by email.Delete a site and all its deploys. It goes offline at once and the name becomes free for anyone. Ask the person first.
| Argument | Type | Description |
|---|---|---|
| site* | string | |
| token | string | A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check. |
| confirm* | boolean | Must be true |
{
"type": "object",
"required": [
"site",
"confirm"
],
"properties": {
"site": {
"type": "string"
},
"token": {
"type": "string",
"description": "A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check."
},
"confirm": {
"type": "boolean",
"description": "Must be true"
}
}
}Publish a static website (HTML, CSS, JS, images) to the Press5 launch pad and get its live URL. Creates the site the first time (the name must be free; 3-42 lowercase letters, digits, hyphens). Send every file of the built site; this replaces the whole site. Text files as plain content, binary files as base64 with encoding "base64". Best for small sites: for big builds use the HTTP API with a .tar.gz (see https://my.press5.xyz/llms.txt).
| Argument | Type | Description |
|---|---|---|
| spa | boolean | Single-page app: serve index.html for unknown paths (client-side routing) |
| site* | string | Site name, becomes <site>.on.press5.xyz |
| files* | array | |
| token | string | A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check. |
{
"type": "object",
"required": [
"site",
"files"
],
"properties": {
"spa": {
"type": "boolean",
"description": "Single-page app: serve index.html for unknown paths (client-side routing)"
},
"site": {
"type": "string",
"description": "Site name, becomes <site>.on.press5.xyz"
},
"files": {
"type": "array",
"items": {
"type": "object",
"required": [
"path",
"content"
],
"properties": {
"path": {
"type": "string",
"description": "Path in the site, e.g. index.html or assets/app.js"
},
"content": {
"type": "string",
"description": "File content: text, or base64 when encoding is \"base64\""
},
"encoding": {
"enum": [
"utf8",
"base64"
],
"type": "string",
"default": "utf8"
}
}
},
"minItems": 1
},
"token": {
"type": "string",
"description": "A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check."
}
}
}List the sites on this Press5 account, with their URLs and status, plus the account limits.
| Argument | Type | Description |
|---|---|---|
| token | string | A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check. |
{
"type": "object",
"properties": {
"token": {
"type": "string",
"description": "A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check."
}
}
}Check whether the person approved the sign-in from login_start. Call it every few seconds after they say they have pressed Approve. Returns the token once.
| Argument | Type | Description |
|---|---|---|
| device_code* | string |
{
"type": "object",
"required": [
"device_code"
],
"properties": {
"device_code": {
"type": "string"
}
}
}Start signing in to the Press5 launch pad when you have no token. Press5 emails the person a link to approve this tool (it also creates a free account for a new email). Then call login_check.
| Argument | Type | Description |
|---|---|---|
| email* | string | The person's email address |
| client | string | Your name, shown to the person, e.g. "Claude Code" |
{
"type": "object",
"required": [
"email"
],
"properties": {
"email": {
"type": "string",
"description": "The person's email address"
},
"client": {
"type": "string",
"description": "Your name, shown to the person, e.g. \"Claude Code\""
}
}
}Put an earlier deploy of a site back live. Get deploy ids from site_info.
| Argument | Type | Description |
|---|---|---|
| site* | string | |
| token | string | A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check. |
| deploy* | string |
{
"type": "object",
"required": [
"site",
"deploy"
],
"properties": {
"site": {
"type": "string"
},
"token": {
"type": "string",
"description": "A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check."
},
"deploy": {
"type": "string"
}
}
}One site with its recent deploys (ids for rollback_site).
| Argument | Type | Description |
|---|---|---|
| site* | string | |
| token | string | A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check. |
{
"type": "object",
"required": [
"site"
],
"properties": {
"site": {
"type": "string"
},
"token": {
"type": "string",
"description": "A press5 token (p5_...). Only needed when the MCP connection has no Authorization header, e.g. right after login_check."
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 19, 2026.
Takes a password, API key or token as an argument, so the secret passes through the model's context and chat logs.
Fix: Read credentials from the server's own configuration (environment, OAuth) instead of a tool argument.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.