Skip to main content
Glama
Pixifi MCP Server

Pixifi MCP Server

README.md
# Pixifi MCP server for ChatGPT and Claude

**ChatGPT, Claude and other AI agents can work with your Pixifi account through Cotlet.dev.** The hosted Pixifi MCP server offers 38 tools for studio paperwork in your own signed-in Cloud Browser. Write tools ask for confirmation by default, including proposals, contracts and invoices that reach a client. Cotlet.dev is independent and is not affiliated with Pixifi.

## What you can ask

> Turn an inquiry into a lead and draft a proposal for my approval.

> Build the wedding day timeline and assign staff.

> Create an event invoice, then ask before sending it.

> Read open invoice balances for this year.

> Read this year's profit and loss statement.

## Connect your AI agent

Sign up at https://cotlet.dev/signup?app=pixifi, enable Pixifi and sign in at https://www.pixifi.com/admin/ in your Cloud Browser.

### ChatGPT

Add `https://cotlet.dev/mcp/v1/connect` in Settings > Apps & Connectors, select OAuth and sign in to Cotlet.dev.

### Claude

Add `https://cotlet.dev/mcp/v1/connect` in Settings > Connectors > Add custom connector and sign in to Cotlet.dev.

### Codex and Claude Code

For Codex, set `url = "https://cotlet.dev/mcp/v1"` and `bearer_token_env_var = "COTLET_API_KEY"` under `[mcp_servers.cotlet]` in your config.
For Claude Code:

```bash
claude mcp add --transport http cotlet https://cotlet.dev/mcp/v1 --header "Authorization: Bearer $COTLET_API_KEY"
```

Get your API key at https://cotlet.dev/app and export it as `COTLET_API_KEY`.

### Cursor

Copy `.cursor/mcp.json` from this repo into your project:

```json
{
  "mcpServers": {
    "cotlet": {
      "url": "https://cotlet.dev/mcp/v1",
      "headers": {
        "Authorization": "Bearer ${COTLET_API_KEY}"
      }
    }
  }
}
```

## Run it locally (stdio)

This repo is a real MCP server. It speaks MCP over stdio, lists its Pixifi tools without any key, and forwards every tool call to the hosted Cotlet.dev server. The hosted URL above stays the recommended way. Use this when your client only starts local servers.

```json
{"mcpServers":{"pixifi":{"command":"uvx","args":["--from","git+https://github.com/cotlet-connect/pixifi-mcp","pixifi-mcp"],"env":{"COTLET_API_KEY":"your-key"}}}}
```

Or with Docker:

```bash
docker build -t pixifi-mcp .
docker run -i --rm -e COTLET_API_KEY pixifi-mcp
```

Tools: `pixifi_guide`, `pixifi_status`, `pixifi_open`, `pixifi_read`, `pixifi_click`, `pixifi_type`, `pixifi_close`. Without `COTLET_API_KEY` a tool call answers: "Set COTLET_API_KEY. Get a key at https://cotlet.dev/app".

## Command line

Python 3.10+, standard library only. Install from this repository:

```bash
pip install git+https://github.com/cotlet-connect/pixifi-mcp
pixifi-mcp connect
```

Or straight from this repo:

```bash
export COTLET_API_KEY="your-key"
python3 pixifi_agent.py connect   # ready config for Claude Code, Cursor, ChatGPT
python3 pixifi_agent.py status    # is Pixifi ready in your Cotlet.dev account
python3 pixifi_agent.py guide     # Pixifi screens and sign-in link
python3 pixifi_agent.py read      # open Pixifi, read the page, close the browser
python3 pixifi_agent.py tools     # list Cotlet.dev tools
python3 pixifi_agent.py serve     # start the stdio MCP server (same as no argument)
```

## Questions from Pixifi users

**Does Pixifi have an API or an MCP server?**

The Cotlet.dev page, checked 16 November 2026, found no public Pixifi API documentation, Pixifi MCP server or official ChatGPT or Claude connector. Pixifi described bespoke API keys for partner vendors in its 1 November 2023 webinar. Cotlet.dev works through your own signed-in Pixifi account.

**How do I connect Pixifi to ChatGPT?**

Sign up at https://cotlet.dev/signup?app=pixifi, enable Pixifi and sign in at https://www.pixifi.com/admin/ in your Cloud Browser. Add https://cotlet.dev/mcp/v1/connect in ChatGPT with OAuth.

**Can I use Pixifi with Claude?**

Yes. Add https://cotlet.dev/mcp/v1/connect as a custom connector in Claude and sign in to Cotlet.dev.

**Does record_refund send money back?**

No. Pixifi records a negative payment of type Refund. You issue the refund yourself in Stripe or Square.

**Which Pixifi plan do the tools use?**

Your own user and plan apply. Pixifi Academy says Essential can assign only yourself to an event, while Workshops require Studio Pro or higher.

## Source

[Pixifi integration and tool reference](https://cotlet.dev/integrations/pixifi-api-mcp-server)

---

More: https://cotlet.dev/integrations/pixifi-api-mcp-server?utm_source=github&utm_medium=readme&utm_campaign=pixifi

Start free: https://cotlet.dev/signup?app=pixifi&utm_source=github&utm_medium=readme&utm_campaign=pixifi

TDQS

A3.7/5.0

Scored across 7 tools

Disambiguation5/5

Each tool has a clearly distinct role in the browser workflow: guidance, readiness check, session opening, page reading, clicking, typing, and closing. There is little risk of confusing one tool for another, especially because read/click/type operate on numbered page elements.

Naming Consistency5/5

All tools use a consistent pixifi_ prefix followed by a short snake_case action or state label. Although guide and status are noun-like rather than pure verbs, the overall naming convention is predictable and easy to scan.

Tool Count5/5

Seven tools is well-scoped for a browser-automation surface: guide and status support setup, open/read/click/type cover the core interaction loop, and close handles cleanup. No tool appears redundant.

Completeness4/5

The set covers the essential browser lifecycle and interaction primitives needed to work inside Pixifi. Minor gaps exist, such as explicit navigation, scrolling, select/dropdown handling, or upload/download support, but most workflows can be worked around via read/click/type.