LocalASOMCP and AI agents
LocalASO competitors_get tool: find the apps ranking ahead of yours
competitors_get returns the apps that rank ahead of yours on your own board keywords, sorted by how often they beat you. It is the quickest way for an agent to answer "who am I losing to, and by how much".
Updated
Input
| Argument | Type | Required | Meaning |
|---|---|---|---|
app | string | yes | Internal id, App Store trackId, or part of the app name |
Output
An object with app (the usual summary) and losingTo, an array. Each entry has:
| Field | Meaning |
|---|---|
trackId | The competitor's App Store ID |
name | The competitor's name |
iconUrl | URL of its icon |
ratingCount | Its rating count |
aheadCount | The number of keyword and country pairs where this app ranks ahead of yours |
aheadCount counts pairs, not keywords. An app with aheadCount: 12 might beat you on 12 keywords in one country, or on 4 keywords in 3 countries. The list is built from the latest checks on the board, so it only reflects keywords and countries that have been checked. Run keywords_check first.
Reading a result
Here is a hypothetical result shape for a habit tracker called Streakly, shown as a table because the values are made up for illustration:
| name | ratingCount | aheadCount |
|---|---|---|
| Habit app A | 480,000 | 31 |
| Habit app B | 90,000 | 17 |
| Habit app C | 2,100 | 4 |
App A has high ratings and beats you everywhere: its position comes from ratings, which you cannot match this quarter. App C has few ratings and still outranks you on four pairs, which suggests a metadata advantage you can learn from. That second kind is where an agent adds value, because you can ask it to compare listings.
Prompts that use it
These are example prompts, not recorded sessions.
- "Which three apps beat me most often? Use competitors_get." The agent reads
losingToand sorts byaheadCount. - "Which of those have fewer than 5,000 ratings?" The agent filters by
ratingCount. - "For the small ones, which of my keywords do they beat me on?" The competitor list does not say which keywords, so a good agent reads board_get for the keywords where your rank is poor and then inspects the live results. If it cannot, it should say so rather than guess.
The tool gives names and counts, never the competitor's keyword field. To compare listings you read public pages yourself or use the App Store keyword field checker on your own draft.
Limits
- Rating counts come from public data and drive LocalASO's volume and difficulty estimates; they are not Apple search counts.
- A tool error such as an ambiguous app name returns as text, and a Mac without a valid license gets a licensing message.
- Competitors found this way depend on your keyword choice. A narrow board gives a narrow rival list.
Next step
Turn the list into action with AI agent prompts for App Store keywords or read the saved text of your own page with listing_get. For the tool overview see the MCP docs.
Questions
- Does competitors_get need checks to have run?
- Yes. It uses the latest checks on the board, so run keywords_check first or the list will be short or empty.
- Does it return other apps' keywords?
- No. It returns the apps and their counts. It does not return their keyword fields.
Do this for every country, in minutes