Free AI app listing grader
This free AI app listing grader checks a ChatGPT plugin, Claude connector or MCP Registry listing against each platform's published limits and review rules, and tells you what to change.
The grader scores each field against the platform's rules. The result table shows each check, the rule it comes from, and what to fix.
One per line.
3 checks to fix on ChatGPT
- Name20/20
- Short description10/10
- Description30/35
- Starter prompts17/20
- Keyword4/15
| Check |
|---|
| App name fits the limit8 of 30 characters. |
| Name avoids platform trademarksNo platform names in it. |
| Subtitle fits the limit27 of 30 characters. |
| Description fits the limit185 of 4000 characters. |
| Description says what the app does31 words. Say what it does, for whom, and two or three tasks it handles. |
| Plain words, no fillerReads plainly. |
| No pricing, promotions or comparisonsNone found. |
| Each prompt is 128 characters or fewerAll fit. |
| 1 to 3 starter prompts3 added. |
| Prompts start with a taskStart these with a verb people say: “Our seamless notes”. |
| Keyword appears where people and models readIn 1 of 4 fields. |
Grade your listing in three steps. The check runs in your browser and takes about a minute.
- 01
Pick a platform.
Choose ChatGPT, Claude or the MCP Registry.
- 02
Paste your listing.
Add the name, short description, description, starter prompts and the main keyword you want to rank for.
- 03
Fix what fails.
Start with red checks. They break a platform rule, so review can reject the listing. Then find the right keywords with the free ASO keyword tool.
The grader checks these rules for each platform.
ChatGPT plugin
| Field | Rule | Source |
|---|---|---|
| displayName | 30 characters or fewer. No generic single words. No "MCP" or "Plugin" suffix | OpenAI submission docs |
| subtitle | 30 characters or fewer | OpenAI submission docs |
| description | 4,000 characters or fewer. No comparisons, pricing or promos | OpenAI submission guidelines |
| starter prompts | Up to 3, each 128 characters or fewer | OpenAI submission docs |
Claude connector
| Field | Rule | Source |
|---|---|---|
| Server name | 100 characters or fewer | Anthropic submission docs |
| One-liner | 200 characters or fewer | Anthropic submission docs |
| Description | 2,000 characters or fewer | Anthropic submission docs |
| Categories | 1 to 5 | Anthropic submission docs |
| Tool names | 64 characters or fewer | Anthropic review criteria |
MCP Registry
| Field | Rule | Source |
|---|---|---|
| name | 3 to 200 characters, reverse-DNS pattern such as io.github.you/server | server.json schema |
| description | 1 to 100 characters | server.json schema |
| title | 1 to 100 characters | server.json schema |
Rules checked 3 October 2026.
Your score is a weighted pass rate from 0 to 100.
A broken hard limit, such as a 35-character ChatGPT name, is a fail. A missed best practice, such as a description that does not name the main task, is a warning.
Frequently asked questions.
Is it free?
Yes. No signup and no card.
Do you store my listing?
No. The grader runs in your browser, so your listing is not sent to gaintage.
Does a high score mean I will rank?
No. It means your listing passes the platform rules. Rank also depends on real use, tool descriptions and competitors.
Where do these rules come from?
From OpenAI's and Anthropic's submission docs and the official MCP server.json schema, checked on 3 October 2026.
Does it check tool descriptions?
No. Use the MCP tool description checker for tool names, descriptions and hints.
New to review? Learn how to submit an app to ChatGPT.