# TOPY.AI · AI CPO

TOPY.AI · AI CPO is an MCP server for Claude Code: a product chief for your coding agent that sorts feedback, holds goals and a decision log, and ranks priorities.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: TOPY-AI-LTD
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/TOPY-AI-LTD/TOPY-AI-CPO-Skills
- Updated: 2026-10-27
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/topyai-ai-cpo

## What it does

Nineteen skills across six plugins sit on top of the remote server, and a bundle pulls all of them in at once. The set covers a weekly product review, a quiet daily check, feedback capture and triage, a survey for product-market fit, RICE ranking, a roadmap, PRDs that name a success metric, experiment planning and a decision log.

Skills come from a catalogue instead of being edited by hand, and the build matches every tool they mention against what the server exposes. Experiments get registered up front: an early read before the planned sample tells you to carry on, not that you have a winner. Each change takes a preview, then a confirmation from you.

You connect to `https://topy.ai/mcp/cpo` over OAuth with a TOPY.AI account. Product records open only for the account that owns the OneMan operation, so any other login leaves the skills empty. The server writes no code and ships nothing; the AI CTO does the building. Pricing, sign-in, permissions and legal copy stay with a person. So does paying for a build or deleting customer data.

## Requirements
- Needs: Node.js (commands run via npx)
- Remote server: `https://topy.ai/mcp/cpo`
- Environment variables: `TOPY_API_KEY` — keys and tokens: ask the human for the values, never commit them or put them in shared files.

## Install

### Claude Code (native platform)

1. **Platform CLI: remote server**
```bash
claude mcp add --transport http cpo https://topy.ai/mcp/cpo
```

2. **Platform CLI**
```bash
claude mcp add --env TOPY_API_KEY=<TOPY_API_KEY> --transport stdio topyai-ai-cpo -- npx -y @topy-ai/cpo-mcp
```
Replace <…> with your own values

### Codex

1. **Platform CLI: remote server**
```bash
codex mcp add topyai-ai-cpo --url https://topy.ai/mcp/cpo
```

2. **Platform CLI**
```bash
codex mcp add topyai-ai-cpo --env TOPY_API_KEY=<TOPY_API_KEY> -- npx -y @topy-ai/cpo-mcp
```
Replace <…> with your own values

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "topyai-ai-cpo": {
            "url": "https://topy.ai/mcp/cpo"
        }
    }
}
```
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "topyai-ai-cpo": {
            "command": "npx",
            "args": [
                "-y",
                "@topy-ai/cpo-mcp"
            ],
            "env": {
                "TOPY_API_KEY": "<TOPY_API_KEY>"
            }
        }
    }
}
```
Replace <…> with your own values

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "topyai-ai-cpo": {
            "type": "http",
            "url": "https://topy.ai/mcp/cpo"
        }
    }
}
```
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "topyai-ai-cpo": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "@topy-ai/cpo-mcp"
            ],
            "env": {
                "TOPY_API_KEY": "<TOPY_API_KEY>"
            }
        }
    }
}
```
Replace <…> with your own values

### Gemini CLI

1. **Platform CLI: remote server**
```bash
gemini mcp add -t http topyai-ai-cpo https://topy.ai/mcp/cpo
```

2. **Platform CLI**
```bash
gemini mcp add -e TOPY_API_KEY=<TOPY_API_KEY> topyai-ai-cpo npx -y @topy-ai/cpo-mcp
```
Replace <…> with your own values

## When to use it
- triage a batch of fresh feedback and answer the people who sent it
- write a PRD with a success metric before engineering starts
- rank upcoming features with RICE
- read an experiment only after the planned sample
- log a decision together with what would prove it wrong

## Example prompts
- Give me the product review for this week
- Triage the new feedback and reply to the senders
- Is this worth building? Run the check before it goes to the AI CTO

## Similar
- [claude-pm-skills](https://affpapa.org/en/nevibecode/skills/claude-pm-skills.md) — skill · 202 installs
- [cpo-review](https://affpapa.org/en/nevibecode/skills/cpo-review.md) — skill · 377 installs
- [discover-product](https://affpapa.org/en/nevibecode/skills/discover-product.md) — skill · 104 installs
- [product-manager-toolkit](https://affpapa.org/en/nevibecode/skills/sickn33-product-manager-toolkit.md) — skill · 3500 installs
- [plaid](https://affpapa.org/en/nevibecode/skills/plaid.md) — skill · 876 installs
- [pm-skills](https://affpapa.org/en/nevibecode/skills/podo-pm-skills.md) — skill · 203 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
