OP3 MCP Server & CLI

What is the OP3 MCP server & CLI?

The OP3 MCP server & CLI is a free, open source program that lets AI agents read open podcast analytics from OP3 for you, in 2 ways. The MCP server is what an AI app like Claude, Codex or Cursor connects to, through MCP (Model Context Protocol), the open standard AI apps use to call outside tools.

You ask in plain language. Your AI picks the right tool, and the server makes the call to OP3's public API.

The CLI is the same program as commands. op3-cli op3-show-downloads --identifier <feed url> runs the same code your AI runs when you ask how your show did this month, whether an agent like Claude Code runs it or you do.

What can you ask it?

Once it's set up, you ask the way you'd ask an assistant. These are real prompts it handles:

Try asking
How many downloads did my podcast get this month?
How many actual people is that, not downloads?
Are my listeners coming back, or is it a new audience every month?
Is this week's episode tracking ahead or behind my usual?
Which podcast apps do my listeners use, and is that unusual?
OP3 shows nothing for my feed. What's wrong?

The second one shows why this is useful. Every podcast dashboard counts downloads, which are apps fetching files, and OP3's rows let this count the people behind them.

How to install the OP3 MCP server

Pick your app in the box at the top of this page. Each one is a single command or a pasted block, with no account and no key to start.

Set up OP3

There's nothing to set up to start. The server falls back to OP3's shared preview token, which is rate limited and which OP3 can withdraw.

You don't need to own a podcast. The token reads public OP3 data, which covers every show with the prefix on its feed.

Check that it works

Run the doctor. It checks the token and makes a live call to OP3.

npx -y @thenavidm/op3-mcp-cli doctor

If OP3 has nothing for a show, the feed has no OP3 prefix yet. op3_verify_prefix checks it.

Use the OP3 CLI

The CLI is the same 22 tools as commands. AI agents that run commands, like Claude Code, Codex and OpenCode, use it on their own, and you can type the same commands in a terminal or a script.

Every tool name becomes a command with dashes, so op3_show_downloads runs as op3-cli op3-show-downloads.

npm install -g @thenavidm/op3-mcp-cli
op3-cli
op3-cli op3-global-app-share
op3-cli op3-get-show --identifier https://feeds.example.com/show.xml
op3-cli op3-audience-summary --help

The bare op3-cli lists every command, and op3-cli <command> --help shows what a command takes. Nothing asks for a confirm check, because nothing here writes.

These flags work on every command:

A script can branch on the exit code:

MCP server or CLI: which one?

Both are the same program with the same 22 tools. The difference is when your AI pays for them, in the tokens it reads.

An app that loads every tool up front sends the full list with every message, whether you mention OP3 or not. Claude Code doesn't by default: its tool search sends only the tool names and the server's instructions, and loads a tool's full definition when your AI reaches for it.

The CLI costs nothing until OP3 comes up. Then your agent reads its skill file once and runs the commands it needs. With the skill added, Claude Code also lists its one-line description with every message, about 190 tokens.

When the whole conversation is about OP3, the gap closes. Then the MCP server is the better experience, because you ask in plain language and your AI never has to look up a command.

How to spend less

In Claude Code, leave tool search on, which it is unless you've set ENABLE_TOOL_SEARCH=false.

Turn the server off when you aren't looking at your stats, with your AI app's own switch.

Or install the CLI and connect the server later, on the days it earns its place. Every tool stays in reach, and the standing cost drops to nothing.

Every number here was measured on September 27, 2026 in Claude Code 2.1.257 with Claude Opus 5. It's the same one-line prompt with and without the server connected, once with tool search off so every tool loads, and once with Claude Code's default. The skill was measured the same way, and other apps and models count tokens a little differently.

Reading the numbers

A few things about podcast numbers explain most of the questions people ask.

A download isn't a person

It's an app fetching a file, and one listener whose app fetches again over 3 days counts 3 times. op3_audience_summary gives both numbers and the ratio.

Compare episodes at the same age

An older episode has had longer to collect downloads. op3_episode_curve compares each one against your show's own median at the same age.

App share needs a benchmark

Nearly every show's biggest app is Apple Podcasts, so 40% on Apple can be below average. op3_benchmark_apps divides by OP3's global share, so 100 means exactly average.

The totals lag about a day

Rolled-up figures carry the date they're as of, which is why they won't match a live dashboard exactly.

Every OP3 tool

All 22 tools, grouped the way the repo groups them. Every one only reads, and the server reaches nothing but OP3.

Shows and episodes

Downloads

Listeners and retention

Trends

Geography

Apps and devices

Discovery

Raw data

Is the OP3 MCP server safe?

There's nothing to undo, because nothing here writes. Every tool reads OP3's public data.

Your listeners stay anonymous too. OP3's per-listener codes are counted inside the server and stripped from everything it returns, so your AI only sees counts, rates and shares.

Your data

There's no server of ours in between. The server runs on your computer, talks to op3.dev and stores nothing on disk.

Answers are kept in memory for 5 minutes, so several questions about one show don't repeat the same slow scan. Set OP3_CACHE_TTL_MS=0 to turn that off.

OP3 MCP server settings

Every setting is optional, as an environment variable in your app's config.

Troubleshooting

Run the doctor first. It names the step that failed and the fix.

If the server doesn't show up in your app at all, run the command your app runs, in a terminal, and read the error.