OP3 MCP
One line: OP3 MCP is a read-only server that answers podcast analytics questions from OP3's data — downloads, unique listeners, geography, apps and growth — plus setup diagnostics.
Find a show: turn a feed URL, podcast:guid or uuid into an OP3 show uuid, look up show details and episode lists.
Download numbers: monthly/30-day and weekly downloads for a show, per-episode counts at 1/3/7/30 days, and side-by-side comparison of up to 20 shows.
Real audience (what OP3's own API can't do): unique listeners vs downloads, new vs returning listeners, cohort retention, and episode audience overlap.
Where and how people listen: geography by continent/country/region/metro/timezone, app share over any window, and a four-dimension device breakdown (agent type, app, device type, device).
Benchmarks: global app market share across all OP3 shows, and a show's app mix indexed against that global mix (100 = average).
Trends over time: downloads and listeners by day/week/month with growth rate, hour-of-day/day-of-week listening patterns, and one episode compared to the show's median at equal age.
Raw data escape hatches: raw download rows with every filter, and the raw request/redirect log for verification.
Discovery and troubleshooting: recent episodes with transcripts, and a prefix check that explains zero-data situations (missing OP3 prefix, prefix not yet seen, or genuinely small numbers).
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@OP3 MCP how many unique listeners did my podcast get last month?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
OP3 MCP
OP3 sits in front of the audio file, so it counts every download whatever app made it.
That is why its numbers differ from Apple's or Spotify's, which each report only their own listeners.
22 tools, including unique listeners, retention cohorts and episode benchmark curves that OP3's own API does not expose.
Built and maintained by Navid Moazzez. Built on Slipway, which turns one definition of each tool into the MCP server and the CLI.
Two ways to use it
Command line
op3-cli runs every tool as a command. 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, a script or a cron job:
npm install -g @thenavidm/op3-mcp-cli
op3-cli # every command
op3-cli op3-global-app-share --agent
op3-cli op3-get-show --identifier https://feeds.example.com/show.xml
op3-cli op3-show-downloads --help
op3-cli which listener retention # find the command for a taskEvery command reads OP3's public data, so nothing asks for --confirm. --agent is JSON on one line for scripts and agents, and --select a,b.c keeps only the fields you name.
op3-cli schema <command> prints the exact JSON Schema an MCP client
receives for that tool.
MCP server, for AI agents
op3-mcp is what Claude Code, Claude Desktop, Cursor and the rest launch.
You never run it by hand:
claude mcp add op3 -- npx -y @thenavidm/op3-mcp-cliIn Claude Desktop, the .mcpb extension
installs on a double click. Section 4 has every other client.
Related MCP server: rybbit-mcp
Contents
1. What you can ask it 💬
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?
Where in the world is my audience, down to the state?
Which podcast apps do my listeners use, and is that unusual?
Compare my show against these three others.
What day and hour do most downloads happen?
Which of my episodes share the same listeners?
OP3 shows nothing for my feed. What is wrong?
The one that is impossible without this server is the second. Every podcast analytics dashboard reports downloads. A download is an app fetching a file, not a person, and one listener whose app re-requests across three days counts three times. OP3's raw rows carry a privacy-preserving per-listener hash, so unique listeners, returning listeners and retention can be computed from them. None of OP3's own aggregated endpoints expose that, and this server does.
2. Quick install ⚡
Node 22 or newer. Nothing else.
npx -y @thenavidm/op3-mcp-cli@latest --versionThat is the whole install. npx fetches it on demand, so there is nothing to
update later: with @latest, a new version reaches you the next time your
client starts the server.
To build from source instead:
git clone https://github.com/thenavidm/op3-mcp-cli.git
cd op3-mcp
npm install
npm run build
npm test
node dist/index.js --version3. Setup 🔑
You can skip this. The server works with no credential at all, because OP3 publishes a shared preview token and that is what it falls back to. It is rate limited and OP3 can withdraw it, so get your own before relying on it.
Have an agent do it
The agent cannot sign in to OP3 for you. What it can do is walk you through it and wire up the config.
Paste this into Claude Code, Cursor, or any agent with terminal access:
Help me set up the OP3 MCP server.
1. Open https://op3.dev/api/keys and tell me what to click to create an
API key and its bearer token.
2. Stop and wait for me to paste the token back.
3. Add the server to my MCP client config with that token as OP3_TOKEN.
4. Run the doctor command and tell me whether it worked.Or do it yourself
Go to op3.dev/api/keys.
Create an API key and copy its bearer token.
Set it as
OP3_TOKENin your client config, as in section 4.
You do not need to own a podcast. The token reads public OP3 data, which covers every show that has the OP3 prefix on its feed.
To revoke
Delete the key at op3.dev/api/keys. It stops working immediately.
If you want your own show in here
Your podcast needs the OP3 prefix on its episode audio URLs. That is a change to your feed, not to this server, and it is documented at op3.dev/setup. Until a listener downloads an episode through the prefix, OP3 has no data for your show and neither does this.
4. Connect your client 🔌
The long version, every step with what to do when one fails, is in INSTALL.md.
OP3_TOKEN is optional in every block below. Leave it out to use OP3's preview
token.
Claude Code
claude mcp add op3 \
-e OP3_TOKEN=your-token \
-- npx -y @thenavidm/op3-mcp-cli@latestAdd --scope user to make it available in every project rather than just this
one.
Claude Desktop
The short way: download the .mcpb extension
from the latest release and double-click it. It carries its own dependencies,
so there is no config file to edit and nothing to install first. It asks for an OP3 token, which is optional.
The long way, if you would rather edit the config yourself:
Platform | Config path |
macOS |
|
Windows |
|
{
"mcpServers": {
"op3": {
"command": "npx",
"args": ["-y", "@thenavidm/op3-mcp-cli@latest"],
"env": { "OP3_TOKEN": "your-token" }
}
}
}Tip Claude Desktop does not inherit your shell PATH. If
npxis not found, use the absolute path fromwhich npx.
Quit Claude Desktop completely and reopen it.
claude.ai on the web
claude.ai runs a connector from Anthropic's cloud rather than your machine, so it cannot launch a local command. It needs a public HTTPS URL.
Run the server over HTTP:
npx -y @thenavidm/op3-mcp-cli@latest --http --port 8000Host it somewhere with a public HTTPS URL, then in claude.ai go to Customize, Connectors, +, Add custom connector, paste the URL and click Add.
On Team and Enterprise plans an owner adds it first under Organization settings, Connectors, then each member enables it under Customize, Connectors.
It binds to 127.0.0.1 by default, which is all a tunnel or a proxy on the
same machine needs. To accept connections from outside the machine, set
OP3_HTTP_HOST=0.0.0.0 together with OP3_HTTP_TOKEN, a bearer token every
request must send: without one, the server refuses to listen beyond this
machine.
Cursor
.cursor/mcp.json, same JSON shape as Claude Desktop, key mcpServers.
Windsurf
~/.codeium/windsurf/mcp_config.json, key mcpServers.
VS Code
.vscode/mcp.json. The key is servers, not mcpServers, and each entry takes
"type": "stdio".
{
"servers": {
"op3": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@thenavidm/op3-mcp-cli@latest"],
"env": { "OP3_TOKEN": "your-token" }
}
}
}Codex CLI
~/.codex/config.toml:
[mcp_servers.op3]
command = "npx"
args = ["-y", "@thenavidm/op3-mcp-cli@latest"]
[mcp_servers.op3.env]
OP3_TOKEN = "your-token"Gemini CLI
~/.gemini/settings.json, key mcpServers.
Everything else
Any stdio MCP client takes the same three things: the command npx, the args,
and the env block.
Docker
docker build -t op3-mcp .
docker run --rm -i -e OP3_TOKEN=your-token op3-mcp5. Check it worked 🩺
npx -y @thenavidm/op3-mcp-cli@latest doctorIt tests the token against the cheapest OP3 endpoint, tests the raw query endpoints separately because they fail for their own reasons, and prints the settings in force.
Symptom | Cause |
| The key was deleted at op3.dev/api/keys, or |
Every show lookup returns not found | The podcast does not have the OP3 prefix on its feed. Run |
Rolled-up tools work, raw ones time out | The window is too wide. Narrow it, or raise |
Numbers do not match the OP3 dashboard | Rolled-up figures lag by about a day. Check the |
| Desktop does not inherit your shell PATH. Use the absolute path |
6. What it costs to have connected
Both surfaces are the same program with the same 22 tools. The difference is when the model pays for them. Measured in Claude Code:
MCP server | CLI | |
Every message, with every tool loaded | 9,000 tokens | nothing |
Every message, Claude Code's default | 1,000 tokens | nothing |
When OP3 comes up | nothing more, or the tools it picks | 3,400 tokens for |
20 messages with OP3 in 1, every tool loaded | 180,000 tokens | 3,400 tokens |
Claude Code's tool search is on by default: it sends only the tool names and the server instructions, and loads a tool's full definition when the model reaches for it. An app that loads every tool up front pays the first line on every message, whether OP3 comes up or not. With the skill added, Claude Code also lists its one-line description, about 190 tokens.
To spend less, turn the server off when you are not using it, which in Claude
Code is the /mcp panel.
Or install the CLI and add the server on the days it earns its place.
Measured on 2026-10-05 against 1.2.2, with Claude Code 2.1.286 on Claude Opus
5.5 (one short prompt with and without the server connected, once with
ENABLE_TOOL_SEARCH=false and once with the default, the difference read from
the API's own usage figures; SKILL.md the same way) and Codex 0.159.3 on
gpt-6.1-sol:
Cost | 1.2.2 | 2.0.0 |
Claude Code, every tool loaded, every message | 9,784 | 9,031 |
Claude Code's default, tool search, every message | 1,030 | 1,026 |
| 3,384 | 3,424 |
Codex over the CLI, one task, median of five | 84,242 | 61,677 |
Codex over MCP, the same task, median of five | 78,381 | 78,341 |
The task was "find the command that compares listener retention between two
periods, and the flags it requires". Every tool loaded costs less because each
tool no longer repeats $schema, additionalProperties and an execution
block. Over the CLI, every 1.2.2 run read the general help, the command list
and the command's help, three requests that each carry the conversation so
far; which answers with the command's help when one command fits well ahead
of the rest, so 4 of five 2.0.0 runs needed two. Over MCP, Codex answered
from the tool list without calling a tool, and read about 40,000 input tokens
in one run of five on each version and about 78,000 in the rest. SKILL.md
costs 40 more because it lists which and exit code 1. Other apps and models
count tokens a little differently, and tool-list characters divided by four
are not API usage.
7. Tools 🛠️
Every tool is a read. Nothing here changes anything.
Finding a show
Tool | Does |
| Any identifier to an OP3 show uuid. Takes a feed URL and encodes it for you |
| Show title, uuid, podcast guid, stats page, optionally the episode list |
| Episodes with OP3 episode ids and publication dates, searchable by title |
Downloads, from OP3's rolled-up data
Fast. Prefer these whenever they answer the question.
Tool | Does |
| Monthly downloads, the week-by-week breakdown, and the weekly average |
| Downloads at 1, 3, 7 and 30 days after publication, plus all-time |
| Several shows ranked side by side, in one request |
Audience, from the raw rows
The tools OP3's own API cannot give you.
Tool | Does |
| Unique listeners against downloads, and the ratio between them |
| First-time against repeat listeners, over a baseline you choose |
| Cohort carry-over between two periods, as two separate rates |
| Which episodes share an audience, and how much |
Where and how
Tool | Does |
| Downloads and listeners by continent, country, region, metro or timezone |
| Podcast apps, over any window rather than OP3's fixed three months |
| All four dimensions at once: agent type, app, device type, device |
| App market share across every show OP3 measures |
| This show against that global mix, indexed so 100 is average |
Over time
Tool | Does |
| Downloads and listeners by day, week or month, with a growth rate |
| Hour of day and day of week, in UTC |
| One episode against the show's median at the same age |
Discovery and setup
Tool | Does |
| Episodes across OP3 that carry a |
| Whether OP3 is receiving downloads for a feed, and what is wrong if not |
| Raw download rows, every filter OP3 offers. The escape hatch |
| Raw request log across OP3. A verification surface, not an analytics one |
8. Reading the numbers 📊
The part worth more than the upstream API docs. All of it was found by probing the live API, because OP3's OpenAPI document declares no response schemas.
A download is not a person. It is an app fetching a file. Podcast analytics
across the industry quotes downloads, and a show with a loyal audience whose
apps re-request looks larger than a show with more actual listeners.
op3_audience_summary gives both numbers and the ratio.
Episodes are not comparable on totals. An older episode has had longer to
accumulate downloads. op3_episode_curve compares at equal age against the
show's own median, and excludes episodes younger than the horizon from that
median so a three-day-old episode does not drag a thirty-day comparison down.
App share on its own says almost nothing. Nearly every show's biggest app is
Apple Podcasts, because Apple is around 38% of all podcast listening. A show at
40% Apple is under-indexed. op3_benchmark_apps divides by the global share
so the number means something.
bots does more than add bots. With it off, OP3 applies its published
download calculation, which deduplicates repeat requests. With it on you get raw
request rows. On one probe the same window returned 213 against 345, and only 79
of the difference were bot rows.
Rolled-up figures lag about a day. They carry an asof date. That is why
they will not match a live dashboard exactly.
An empty episode list is normal for a small show. OP3's daily rollup has not covered it yet. The show-level total may still be there.
Metro codes are US-only. They are a US broadcast concept. Use
level=region for a worldwide breakdown.
Raw download rows come back oldest first. OP3 offers no way to reverse that on this endpoint, so a limit takes the earliest rows in the window. Narrow the window to see recent activity rather than raising the limit.
Some shows cannot be filtered by episode. OP3 derives the episode id from
the episode audio URL, so a host that regenerates those URLs leaves historical
rows carrying ids that no longer appear in the feed. Checked across three shows,
two matched exactly and one had no overlap at all. op3_episode_curve detects
this and says so rather than returning silent zeros.
The raw endpoints are scans, not indexes. Two rows took 2790ms when probed. Cost grows with the window. The server pages with a continuation token, caps what it will pull, and labels any result that was cut short, because a truncated result presented as complete makes every rate computed from it wrong.
9. Your data 🔐
There is no backend. The server runs on your machine, talks to op3.dev, and
stores nothing on disk.
Responses are cached in memory for five minutes so an agent asking several
questions about one show does not pay for the same scan repeatedly. That cache
dies with the process. Set OP3_CACHE_TTL_MS=0 to disable it.
Per-listener identifiers never leave the server. OP3's raw row carries
audienceId and hashedIpAddress. Both are aggregated over inside the process
and stripped from anything returned. Every audience figure is a count, a rate or
a distribution.
This is deliberate. OP3 exists to be a privacy-preserving analytics service, and
a wrapper that streamed stable per-listener keys into a model context would undo
that. op3_query_downloads can emit a shortened, non-reversible label if you
need to tell rows apart by listener, and it is off by default.
Third-party text. Show and episode titles come from arbitrary RSS feeds. Anyone can publish a podcast, so that text is attacker-controlled and reaches the model inside a tool result. The server neutralises fence-breaking and tells the model in its instructions to treat it as data. That raises the cost of an injection rather than removing it. The real reason the blast radius is small is that this server is read-only and reaches nothing but OP3.
Settings
Variable | Default | Does |
| preview token | Your bearer token from op3.dev/api/keys |
| unset | Accepted as an alias for |
| 45000 | Per-request deadline |
| 150 | Spacing between requests |
| 3 | Retries on 429 and 5xx |
| 50000 | Cap on rows any one analysis pulls |
| 40 | Cap on continuation pages |
| 300000 | Response cache lifetime, 0 disables |
| op3-mcp | Sent on every request |
| OP3 API root, leave alone unless you proxy it | |
| 8787 | Port for |
| 127.0.0.1 | Bind address for |
| none | Require this bearer token on HTTP requests |
| none | Comma-separated browser origins allowed to call |
| full |
|
| none | Give up on any tool after this long |
| 0 |
|
10. Troubleshooting 🔧
Run doctor first. It names the problem in one command.
npx -y @thenavidm/op3-mcp-cli@latest doctorSymptom | Cause and fix |
"OP3 has nothing at /shows/..." | The show has no OP3 prefix on its feed. Run |
Downloads are zero but the show exists | The prefix was added recently and nothing has come through. Widen the lookback |
Every row is a bot | The prefix works but no listener has downloaded yet |
A tool times out | The window is too wide for a scan. Narrow it or raise |
A result says it was truncated | A cap stopped the pull. Narrow the window, or raise |
| The host regenerates episode URLs, so episode ids do not line up. Show-level tools still work |
Rate limited on the preview token | It is shared. Get your own at op3.dev/api/keys |
FAQ ❓
An MCP server is a standard way to give an AI assistant tools it can actually call. Model Context Protocol is the agreement they speak, so any MCP client connects to any MCP server. This one exposes 22 read-only tools over OP3.
op3-cli is the same program as the MCP server, run 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, a script or a cron job. Every tool is a command with dashes, so op3_show_downloads runs as op3-cli op3-show-downloads.
Use the MCP server in an app with no terminal, like Claude Desktop's chat. Use the CLI anywhere commands run: an agent like Claude Code, Codex or OpenCode, a script or a cron job. The MCP server's tools take up context on every message, and the CLI costs nothing until it runs.
It does not. The token reads public OP3 data, which covers every show that has the prefix on its feed.
You do not need an account, though you should get a token. Without one the server uses OP3's shared preview token, which is rate limited.
It works with any podcast whose feed carries the OP3 prefix, whoever hosts it.
Those report only their own listeners. OP3 sits in front of the audio file, so it sees every download regardless of app.
It does not. OP3's API is read-only and so is this.
It does not. Per-listener identifiers are aggregated inside the server and stripped from every response.
The raw endpoints are scans. Cost grows with the window. The rolled-up tools answer in milliseconds; prefer them.
It works over the HTTP transport, which needs a public HTTPS URL. See section 4.
With @latest in the install line, the next published version reaches you the next time your client starts the server.
Questions
Run into a problem or have a question? Open an issue and I will help.
About the author 👋
Navid Moazzez is a leading AI business strategist, and the host of the AI Creator Summit, watched by 100,000+ creators. He helps creators and founders master AI and build their own AI Operating System (AI OS) to automate their business and life. He creates useful free tools, MCP servers and CLIs that creators and founders can use in their own workflows.
Links
Personal website: navid.me
Navid Media: navid.media
YouTube: @thenavidm and @thenavidai
X: @thenavidm
Instagram: @thenavidm
LinkedIn: thenavidm
If this is useful, star the repo and come say hi on X.
Dependencies
Library | License | What it does |
Apache-2.0 | The MCP server and the CLI from one definition of each tool | |
Apache-2.0 | The MCP protocol and its transports, through Slipway | |
MIT | Tool argument schemas and validation |
Nothing else. The OP3 client, the pagination, the aggregation and the time handling are all built in.
License
MIT. Free to use, modify, and share.
Not affiliated with, endorsed by, or connected to the Open Podcast Prefix Project.
© 2026 Navid Media. Made with ❤️ by Navid Moazzez.
Available Tools
22 toolsop3_audience_summaryAudience summaryARead-onlyIdempotent
Unique listeners against downloads for a show over any window. This is the number OP3's own dashboard and every rolled-up endpoint cannot give you: how many people, not how many requests. Read downloadsPerListener as the key figure, near 1 means each download is a distinct person, well above 1 means apps re-requesting the same file and every download number you see elsewhere is inflated by that factor.
Inputs: identifier (required), end, bots, start.
op3_benchmark_appsBenchmark appsBRead-onlyIdempotent
A show's app mix against OP3's global mix, with an index where 100 means exactly average. This is the tool that turns app share into something actionable. A show can be 40% Apple Podcasts and be under-indexed, because Apple is around 38% globally, so raw share hides the real story. Over-indexed apps are where this audience is unusual and where it can be reached deliberately.
Inputs: identifier (required), end, bots, start, +1 more.
op3_compare_showsCompare showsARead-onlyIdempotent
Compare several shows side by side on monthly downloads and weekly average, ranked. OP3 accepts many shows in one request, so this is one call rather than several. Useful for benchmarking a show against others in its category, or for tracking a portfolio of shows at once.
Inputs: identifiers (required).
op3_device_breakdownDevice breakdownARead-onlyIdempotent
How a show is consumed, across all four dimensions OP3 records: agentType (app, browser, bot), agentName (the specific app), deviceType (mobile, computer, tablet, smart speaker) and deviceName (Apple iPhone, Android Phone). One call returns all four, which is what you want for a picture of the audience rather than a single ranked list.
Inputs: identifier (required), end, top, bots, +1 more.
op3_download_trendDownload trendARead-onlyIdempotent
A show's downloads and unique listeners over time, bucketed by day, week or month, with a growth rate and the peak period. Growth compares the two halves of the window rather than first period against last, because podcast downloads are weekly-seasonal enough that comparing endpoints is close to noise.
Inputs: identifier (required), end, bots, start, +1 more.
op3_episode_curveEpisode curveARead-onlyIdempotent
How one episode is tracking against the show's own median at the same age. Returns cumulative downloads by day after publication alongside the median across comparable episodes, plus a verdict. This is the only fair way to judge a recent episode: total downloads always favour older episodes because they have had longer to accumulate. Only episodes at least as old as the horizon go into the median, so a three-day-old episode does not drag a thirty-day comparison toward zero.
Inputs: episode_id (required), identifier (required), bots, horizon_days, +1 more.
op3_episode_downloadsPer-episode download counts for a show's recent episodesARead-onlyIdempotent
Per-episode download counts for a show's recent episodes: downloads in the first 1, 3, 7 and 30 days after publication, plus all-time. Use it to compare how episodes performed at equal age. Note that an empty result is normal for a small or new show, it means OP3's daily rollup has not covered it yet rather than that there is no data.
Inputs: identifier (required), limit.
op3_episode_overlapWhich episodes share an audienceARead-onlyIdempotent
Which episodes share an audience. For each pair of a show's busiest episodes, reports shared listeners, a symmetric similarity, and the share of the smaller episode's audience that also heard the other. Read the smaller side's share first: for one big episode and one small one the symmetric number is always near zero, while the smaller side's share tells you whether a spike brought genuinely new people or just gave the existing audience another download.
Inputs: identifier (required), end, bots, start, +1 more.
op3_geographyGeographyARead-onlyIdempotent
Where a show's downloads come from, at whichever level you ask for: continent, country, region (state or province), metro area, or timezone. Reports downloads and unique listeners side by side for each place, because one enthusiastic listener in a small market otherwise looks like a market. Region and metro are qualified by country, so Illinois and Israel do not merge.
Inputs: identifier (required), end, top, bots, +2 more.
op3_get_showLook up a show on OP3ARead-onlyIdempotent
Look up a show on OP3: its uuid, title, podcast guid and public stats page. Set include_episodes to also get the episode list with OP3 episode ids, titles and publication dates. Episode ids from here are what the episode-level tools filter on.
Inputs: identifier (required), episode_limit, include_episodes.
op3_listener_retentionCohort carry-over between two periodsARead-onlyIdempotent
Cohort carry-over between two periods: how much of the earlier period's audience showed up again in the later one. Returns two rates that answer different questions. retentionRate is the share of the old audience that came back, which measures whether the show holds people. carryOverShare is the share of the new period that is old faces, which measures whether the show is growing or recycling. A show can score well on one and badly on the other.
Inputs: identifier (required), bots, later_end, cohort_end, +1 more.
op3_listening_patternsListening patternsARead-onlyIdempotent
When downloads happen, by hour of day and day of week, in UTC. Useful for choosing a publication slot. Read it as request timing rather than listening behavior: a podcast app's scheduled background refresh fires on the app's schedule, not when a person pressed play, so the peaks partly reflect app defaults.
Inputs: identifier (required), end, bots, start.
op3_list_episodesList episodesARead-onlyIdempotent
List a show's episodes with their OP3 episode ids, titles and publication dates, newest first. Use it to find the episode id for a specific episode before asking anything episode-level. Episode titles come from the publisher's RSS feed and are third-party text.
Inputs: identifier (required), limit, search.
op3_new_vs_returningNew vs returningARead-onlyIdempotent
Split a show's listeners in a window into first-time and returning, by comparing against a longer baseline period immediately before it. Answers whether a show is reaching new people or serving the same audience repeatedly. The baseline length matters: a short baseline calls a monthly listener new, so it defaults to four times the window and both sizes are reported so you can judge the answer.
Inputs: identifier (required), end, bots, start, +1 more.
op3_query_downloadsRaw download rows for a show, with every filter OP3 offersARead-onlyIdempotent
Raw download rows for a show, with every filter OP3 offers. This is the escape hatch for questions the aggregating tools do not cover. Prefer op3_show_downloads, op3_audience_summary, op3_geography or op3_app_share when they can answer the question, because raw rows are slow to fetch and expensive to reason over. Rows come back oldest first and OP3 offers no way to reverse that on this endpoint, so the limit takes the earliest rows in the window: to see recent activity, narrow the window with start rather than raising the limit. Per-listener identifiers are removed from the output.
Inputs: identifier (required), end, bots, limit, +3 more.
op3_query_hitsQuery hitsARead-onlyIdempotent
Raw request rows from OP3's redirect log, across every show unless filtered by url. This is the lowest level OP3 exposes and it is a verification surface rather than an analytics one: it shows individual requests including user agents, byte ranges and the edge that served them. For questions about a specific show's performance, use the show tools instead.
Inputs: end, url, limit, start, +1 more.
op3_recent_transcriptsRecent transcriptsARead-onlyIdempotent
Recently published episodes across all of OP3 that carry a podcast:transcript tag, newest first. Use it to find podcasts whose episodes are machine-readable, which is the shortlist worth building on for anything that needs to read transcripts rather than count downloads. Returns podcast guids you can pass straight to the show tools.
Inputs: limit, resolve_shows.
op3_resolve_showTurn any podcast identifier into an OP3 show uuidARead-onlyIdempotent
Turn any podcast identifier into an OP3 show uuid. Accepts a show uuid, a podcast:guid, or an RSS feed URL, and reports which kind it recognized. Use this first when you have a feed URL and need the uuid every other tool wants. If this fails, the show most likely does not have the OP3 prefix on its feed, which means OP3 has no data for it at all.
Inputs: identifier (required).
op3_show_downloadsA show's headline download numbersARead-onlyIdempotent
A show's headline download numbers: downloads in the last 30 days, the week-by-week breakdown over the last four weeks, and the weekly average. This is the fast answer to 'how many downloads does my show get' and should be preferred over the raw query tools whenever it can answer the question. Excludes bots. Updated once a day.
Inputs: identifier (required).
op3_verify_prefixVerify prefixARead-onlyIdempotent
Check whether OP3 is actually receiving downloads for a show, and say what is wrong when it is not. Run this first whenever the numbers are zero or a show cannot be found. It separates the three cases that look identical from the outside: the prefix was never added to the feed, it was added but no download has come through yet, or it is working and the answer is genuinely a small number.
Inputs: identifier (required), lookback.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 22 tools
Each tool targets a distinct analytical dimension (downloads, unique listeners, geography, apps, devices, retention, overlap, raw rows), and the descriptions explicitly say when to prefer one over another. The few superficially similar tools (op3_show_downloads vs op3_query_downloads vs op3_episode_downloads, or op3_app_share vs op3_benchmark_apps vs op3_global_app_share) are clearly differentiated by scope and intent.
All 22 tools use a consistent snake_case verb_noun pattern uniformly prefixed with op3_. No camelCase, no mixed conventions, and the names read predictably.
22 tools is on the heavier side but justified by a genuinely broad analytics domain spanning downloads, listeners, geography, apps, devices, retention, and raw verification surfaces. Each tool earns its place rather than being redundant.
The surface covers the full analytic lifecycle: identifier resolution, show/episode metadata, aggregate metrics, audience composition, breakdowns, trends, cohort retention, raw escape hatches, and a diagnostic verification tool. No obvious dead ends for the stated podcast-analytics purpose.
Maintenance
Related MCP Connectors
Query your Betterlytics web analytics from AI agents: traffic, funnels, journeys, errors, uptime.
GA4, Google Ads and Search Console in Claude. Read-only OAuth, multi-account for agencies.
Connect your ads, shop, analytics, social, CRM and finance platforms once, then let Claude, ChatGPT, Cursor or any MCP client read, join and explain your numbers. Public statistics from the World Bank, IMF, Eurostat, OECD, WHO and SEC filings come as context, searchable and chartable from the same tools. Read-only by design, every number carries its source.
Read-only analytics for Convex apps, queryable via MCP from Claude, Cursor, and other clients.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceProvides statistical analysis and usage tracking for Claude Code, enabling users to query daily/monthly stats, compare users, analyze trends, detect anomalies, and generate usage reports through natural language.1-
- AlicenseBqualityBmaintenanceEnables querying Rybbit Analytics data directly through MCP-compatible clients like Claude Code. It provides tools for monitoring website statistics, user sessions, error logs, funnels, and performance metrics via natural language.4314 npm4MIT
- FlicenseAqualityDmaintenanceEnables natural language querying of Google Analytics 4, Google Search Console, Meta Ads, and Google Ads data through Claude.23-
- AlicenseNot gradedqualityBmaintenanceEnables querying and analyzing marketing data from Google Analytics 4, Search Console, YouTube, and optionally Google Ads and Meta Ads, stored in a unified DuckDB database, via natural language in Claude.1MIT