How to connect Chatmu in Claude
Setup is short. Open Claude, go to Connectors, search for "Chatmu" and click Add. From there, Claude can look up your artist's data through Chatmu. That is all you need for the three prompts below. If you want to understand what an MCP connector is and why it matters, read the complete guide to MCP for the music industry. It also explains the idea behind a music manager Claude connector.
In every prompt, replace [ARTIST] with the name of the artist you manage.
Prompt 1: a full artist report as a PDF
This is the simplest and the biggest time saver: a report ready to share with the artist, covering the metrics they usually ask about in a meeting. It asks for the charts at the end so the text reads first.
Give me a complete report on [ARTIST] as a PDF, ready to send to the artist. Include monthly listeners, demographics, top cities, growth over the last 4 weeks, top songs, active playlists and a comparison across platforms. Put the charts at the end and deliver it to me as a PDF.
Review the result before sending it. If you want to benchmark your artist against others, we have a guide on how to compare artists' streaming performance with AI.
Prompt 2: where are shows viable and how many tickets would sell?
The usual industry rule says that 2% of an artist's monthly listeners in a city will buy a ticket. It is a starting point, not an answer. This prompt adjusts it using the artist's growth, past shows, the date and the market; and if you have your previous sales, the estimate gets more precise. Paste it in full:
I want to know in which cities it is viable for [ARTIST] to play shows and how many tickets they could sell in each one. Use the Chatmu MCP. If you have the chatmu-show-viability skill, load it and follow it.
If you don't have it, do this:
1. Starting point (2% rule): with artist_top_geographic_data take their top Spotify cities. In each city, 2% of monthly listeners is the base attendance. Example: 10,000 listeners in Los Angeles = 200 people. If it comes to 100 or more, a paid show is viable; if it is under 100 (fewer than 5,000 listeners), a free show or opening for another artist is better.
2. Ask me ONLY ONCE whether I have results from past shows (city, date, capacity, tickets sold, price and how many tickets each person buys). If I give them to you, calculate my artist's real conversion (tickets sold ÷ listeners in that city) and use it instead of the 2%: the more shows in a city, the more weight you give that data. If I don't have them, continue with the rest.
3. Adjust each city with what you can check in Chatmu and explain each adjustment:
- Listener growth by city (geographic_growth_analysis): if it is growing, raise the estimate.
- Past shows (get_artist_live_history): if they played in that city less than 6 months ago, lower it; if more than a year ago, raise it.
- Audience age (audience_demographics): if very young, watch out for 21+ venues in the US.
- Reliance on a single song or on playlists (get_artist_songs, get_artist_active_playlists): if almost everything comes from one hit, lower it.
- Date: January, Christmas, Holy Week and local holidays lower it; Friday, Saturday, spring and September through November raise it.
- Market: in their home country or in US cities with a large Latin community (if they are a Latin artist), raise it.4. Deliver to me as a PDF a prioritized list of cities with: monthly listeners, base attendance (2%), realistic range (conservative / realistic / optimistic), confidence level, recommended venue size, whether it makes sense to charge or play for free to grow the base, and why.
5. At the end, tell me which data you are missing so the prediction can be more precise.
Why this prompt works
Three ideas make it useful:
- It does not stop at 2%: it adjusts the figure with real Chatmu data such as growth by city, show history and audience age.
- It asks about your past sales. If you have them, the estimate uses your own artist's real conversion rather than the general rule.
- It gives you a range and the reason behind each number, not a magic number. That makes it easier to discuss with the artist or the promoter.
The technical names (artist_top_geographic_data, get_artist_live_history and the rest) are connector tools; leave them as they are. The chatmu-show-viability skill is optional: the prompt includes the steps in case you do not have it.
Prompt 3: how did the artist do on their last 3 Instagram posts?
A quick review of recent content, useful before a meeting with the artist or a brand.
How did [ARTIST] do on their last 3 Instagram posts? Use the Chatmu MCP to review their recent content and measure the engagement of each post: likes, comments, reach and what worked best.
How to adapt them and where the limits are
- They are estimates. The report and the ticket ranges are estimates based on data and a general rule. Verify the key figures before using them in a negotiation.
- If your Claude plan does not produce PDFs, ask for a table or structured text and move it into your own document.
- Change the scope. Adjust the period, country, platforms or number of posts to what you need.
- The decision is human. Prompt 2 gives you a range and reasons; you decide where to play, what to charge and with which promoter.
- Review before sending. Any report that goes to the artist should pass through you first.
Chatmu is an AI agent for the music industry with its own catalog of music-industry MCP connectors. If you manage several artists, take a look at Chatmu for teams. To see how this fits into your daily operation, read AI agents for artist managers and the comparison of AI music manager apps.
FAQ
What are AI prompts for music managers?
They are ready-to-paste instructions for an assistant such as Claude so that it looks up an artist's data and returns a report, an estimate or an analysis. Here there are three: a full report, show viability and an Instagram review.
Do I need Chatmu to use them?
The prompts are designed for the Chatmu connector in Claude, which is what gives them access to the artist's data. Without a connector that has real data, the assistant can only answer from general knowledge.
How reliable is the ticket estimate?
It is an estimate, not a promise. It starts from the 2% rule as a common reference and adjusts it with real data. The more past sales you provide, the closer it gets to your case. Validate the result before committing dates or budget.
Does an AI music agent generate songs?
No. An AI agent for the music industry works on information and workflows, such as reports, market analysis and content tracking. It does not create audio.
Can I send the report to the artist without reviewing it?
We do not recommend it. Check the figures, tone and conclusions before sending anything. If your Claude plan does not produce PDFs, ask for a table or text and build the document yourself.
AI for the industry. Humans for the music.
