GitHub
GitHub’s official server lets AI read and manage code, issues and pull requests in your account. Use GitHub’s hosted version or run it locally with Docker.
Description from the maker (original English)
Dev-registry data: npm/PyPI/Docker/VS Code packages, dep graphs, vulns, 50+ ecosystems.
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
You need
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Works in Claude on the web, desktop and mobile:
https://devstack-mcp.vercel.app/mcpRun this in your terminal
claude mcp add --transport http devstack https://devstack-mcp.vercel.app/mcpPaste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json
{
"mcpServers": {
"devstack": {
"url": "https://devstack-mcp.vercel.app/mcp"
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Paste into .vscode/mcp.json in your project
{
"servers": {
"devstack": {
"type": "http",
"url": "https://devstack-mcp.vercel.app/mcp"
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
ChatGPT connects only to servers on the internet:
https://devstack-mcp.vercel.app/mcpMenu names can differ slightly by plan and over time.
Your choice stays on this device in a cookie; no account needed.
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
We have not read this server’s tool list yet; we check endpoints in rotation.
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 6, 2026
The registry name was verified with the GitHub account IsaiahDupree, which owns the repo
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
Not checked yet; we check endpoints in rotation
Not checked yet; we check endpoints in rotation
Not checked yet; we read the tool list when we check the endpoint
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 1 of 1 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
The registry does not list a required key, though some servers ask you to sign in on first use.
It runs in the maker’s cloud. Connect over the internet with nothing to install, but what you send passes through the maker’s servers.
The GitHub repo last changed –, and the latest version (1.0.0) reached the registry 3 months ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. It connects by URL, so apps that accept remote MCP servers work too, such as ChatGPT in developer mode.
In the same category: GitHub, Desktop Commander, Cloudflare MCP
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
GitHub’s official server lets AI read and manage code, issues and pull requests in your account. Use GitHub’s hosted version or run it locally with Docker.
MCP server for terminal commands, file operations, and process management
Cloudflare MCP servers
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
A remote MCP (Model Context Protocol) connector for DevStack API — the libraries.io successor: one keyless, normalized view of the whole developer registry across npm, PyPI, Docker Hub, and the VS Code Marketplace (search, package details, versions, download/pull stats), plus resolved dependency graphs (deps.dev), per-version vulnerabilities (OSV.dev), OSSF Scorecard insights (deps.dev), and cross-registry lookups over 50+ ecosystems (ecosyste.ms).
Target deploy: https://devstack-mcp.vercel.app/mcp — 10 tools. Since the upstream DevStack deployment is metered (RapidAPI/Apify) and gates /v1/* behind a RapidAPI proxy-secret guard, this connector authenticates its own outbound calls with that same proxy secret (DEVSTACK_MCP_PROXY_SECRET, sent as the X-RapidAPI-Proxy-Secret header) and applies a soft per-IP rate limit (DEVSTACK_MCP_RATE_LIMIT, default 30 tool-calls/hour, in-memory) so the free MCP tier stays a discovery channel rather than an unmetered bypass of the paid listing — see lib/ratelimit.js.
DevStack API is a plain REST API. Any HTTP client can already call it directly. This repo exists because MCP clients (Claude, ChatGPT, and other MCP-aware agents) don't consume arbitrary REST APIs — they consume MCP tools. devstack-mcp is a thin adapter layer that:
/mcp endpoint, so it can be registered as a connector in Claude, ChatGPT, or any other MCP client with one URL.fetch to DevStack, and the JSON response DevStack returns is handed back verbatim as the tool result.DevStack's data has no per-user dimension at all — it's public developer-registry metadata (package listings, versions, download counts, dependency graphs, published CVEs). There is nothing to gate per-caller, so this connector intentionally ships with:
The one server-side secret it carries (DEVSTACK_MCP_PROXY_SECRET) is not per-caller auth — it's how this connector, as a single consumer, gets past the upstream's RapidAPI proxy-secret guard to reach real data. Combined with the soft per-IP rate limit, that keeps this a free discovery tier over the metered DevStack product rather than an unmetered bypass of the paid RapidAPI/Apify listing.
One tool per DevStack /v1 endpoint (from devstack-api/openapi.yaml):
| Tool | DevStack endpoint | Description |
|---|---|---|
get_package | GET /v1/package | Normalized package details by registry + name (registry=all fans out). |
search_packages | GET /v1/search | Search npm/Docker/VS Code (or all); PyPI has no search API. |
get_versions | GET /v1/versions | Published versions / image tags / extension versions. |
get_downloads | GET /v1/downloads | Download / pull statistics (npm + PyPI only). |
get_dependency_graph | GET /v1/deps | Fully resolved dependency graph for one exact version (deps.dev). |
get_vulnerabilities | GET /v1/vulns | Known CVEs/GHSAs per package/version (OSV.dev). |
scan_vulnerabilities_batch | POST /v1/vulns | Batch/lockfile vulnerability scan — up to 100 queries in one call (OSV.dev). |
get_insights | GET /v1/insights | OSSF Scorecard + repo signal + licenses/advisories (deps.dev). |
search_ecosystems | GET /v1/ecosystems/search | Cross-registry exact-name lookup across 50+ ecosystems (ecosyste.ms). |
get_ecosystems_package | GET /v1/ecosystems/package | One package on one registry, with reverse-dependency counts (ecosyste.ms). |
All 10 tools are read-only and annotated { readOnlyHint: true, destructiveHint: false, idempotentHint: true, openWorldHint: true }. scan_vulnerabilities_batch is an HTTP POST but is a read-only *query* (it mutates nothing), so it carries the same read-only annotations.
Each tool handler does a plain fetch(\${DEVSTACK_MCP_API_BASE_URL}${path}\, ...) against the real DevStack REST API and returns the parsed JSON as MCP tool-result content:
{ content: [{ type: 'text', text: JSON.stringify(result, null, 2) }] }
Upstream HTTP errors (4xx/5xx) are caught and surfaced as a typed MCP error result via an asError helper rather than crashing the request.
| Env var | Purpose | Default |
|---|---|---|
DEVSTACK_MCP_API_BASE_URL | Upstream DevStack base URL. | https://devstack-api.vercel.app |
DEVSTACK_MCP_PROXY_SECRET | RapidAPI proxy secret, sent as X-RapidAPI-Proxy-Secret to pass the upstream /v1/* guard. | "" (omitted) |
DEVSTACK_MCP_RATE_LIMIT | Soft per-IP tool-calls/hour cap for this free tier. | 30 |
api/mcp.js MCP endpoint (StreamableHTTPServerTransport, stateless, no per-caller auth)
api/health.js GET /api/health lib/tools.js All 10 tool definitions (zod schemas + fetch-and-forward handlers) lib/ratelimit.js Soft per-IP tools/call cap (free discovery tier) local-server.js Plain-Node http server for local dev / smoke testing (not deployed) test/smoke.mjs Real end-to-end smoke test (initialize, tools/list, tools/call) vercel.json Routes /mcp -> api/mcp.js, /health -> api/health.js server.json MCP registry publish manifest
npm install
npm run dev # starts local-server.js on http://localhost:3900
Endpoints locally: POST http://localhost:3900/mcp, GET http://localhost:3900/health.
To point at a self-hosted / open DevStack instance instead of production (useful for a fully keyless end-to-end run, since production's /v1/* is behind the RapidAPI proxy-secret guard):
# In devstack-api (guard is inert when RAPIDAPI_PROXY_SECRET is unset):
node server.js # listens on :8080
DEVSTACK_MCP_API_BASE_URL=http://localhost:8080 npm run dev
npm run dev & # terminal 1
npm run smoke # terminal 2
test/smoke.mjs drives the running server over real HTTP/JSON-RPC and verifies:
GET /health returns { ok: true, ... }initialize succeeds and reports serverInfo.name === "devstack"tools/list returns all 10 tools with their zod-derived JSON schemastools/call for get_package with { registry: "npm", name: "react" } returns real package data fetched live (not mocked)Against production, step 4 needs DEVSTACK_MCP_PROXY_SECRET set (the RapidAPI proxy guard); against a locally self-hosted devstack-api with no RAPIDAPI_PROXY_SECRET, the guard is inert and real data flows keyless.
Not deployed by this build. Once ready:
cd /Users/isaiahdupree/Software/devstack-mcp
npx vercel --yes --prod
After deploy, set DEVSTACK_MCP_PROXY_SECRET in the Vercel project so the connector can reach the guarded production /v1/*, then register the MCP connector URL in Claude/ChatGPT/any MCP client:
https://<deployment-domain>/mcp