cd /news/developer-tools/the-mcp-server-pre-publish-checklist · home topics developer-tools article
[ARTICLE · art-27719] src=dev.to ↗ pub= topic=developer-tools verified=true sentiment=· neutral

The MCP Server Pre-Publish Checklist

Incultnitollc released mcp-probe, an open-source CLI tool that scores MCP servers on a 0-100 publishability scale across five axes. The tool enforces a 10-point checklist addressing common failures such as thin tool descriptions, naming collisions, and missing install metadata. Official Anthropic reference servers score only 60/100 on description quality, while typical first-draft community servers land in the 40s.

read2 min publishedJun 15, 2026

Before you publish an MCP server, run 10 checks. Most servers fail at least three — and the failures are invisible until an agent picks the wrong tool, hallucinates an argument, or silently drops your server on connect. This is the checklist we built mcp-probe

to enforce, distilled to what actually breaks in the wild.

TL;DR — A publishable MCP server connects cleanly, names tools unambiguously, describes every argument, validates inputs, and ships install metadata. The single most common failure is thin tool descriptions: even the five official Anthropic reference servers cap at

60/100on description quality.

MCP Inspector answers "does my server connect and list tools?" That's necessary, not sufficient. The agent doesn't experience your server the way you do in a UI — it experiences your tool descriptions and schemas as text in a context window, and it picks tools by reading them. A server can pass Inspector and still be functionally unpublishable because the model can't tell your tools apart.

So the pre-publish question isn't "does it run?" It's "is it publishable?" — will a real agent, with no docs and no human in the loop, use it correctly?

initialize

.initialize

fast.create_issue

exists in a dozen servers. If yours collides, the model guesses. Namespace or specify.server.json

so the official MCP Registry can discover you. Devs find tools at install-time, not search-time.You can walk this list by hand, or run it:

npx @incultnitollc/mcp-probe score "node ./your-server.js"

mcp-probe

connects to your server, runs all ten checks, and returns a 0–100 publishability score across five axes — description quality, enum/shape correctness, mutation legibility, anti-"restate the name" clauses, and distribution metadata. A passing server clears ~80. The official reference servers sit at 60 (the description cap fires on every one). A typical first-draft community server lands in the 40s.

Wire it into CI so it runs on every release:

- run: npx @incultnitollc/mcp-probe score "node ./dist/server.js" --fail-under 80

The exit code gates the publish. Your server can't regress below the bar you set.

If you do nothing else: rewrite your tool descriptions so a model with no context could choose correctly between yours and a similarly named tool. That single fix moves more servers across the publishable line than any other on this list — and it's the one almost everyone skips.

mcp-probe

is an open-source CLI for testing and scoring MCP servers before you publish. npx @incultnitollc/mcp-probe

· github.com/incultnitollc/mcp-probe

── more in #developer-tools 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/the-mcp-server-pre-p…] indexed:0 read:2min 2026-06-15 ·