GitX MCPTry GitX MCP
GitX MCP

GitHub momentum, exposed to AI agents.

GitX MCP is an MCP-native intelligence layer that lets AI coding tools analyze whether GitHub repositories are active, abandoned, stable, or gaining real developer momentum. Built for developers who need signal, not noise.

Try DashboardRead Docs

Analyze a repository

What GitX MCP does

Most GitHub metrics tell you what happened in the past. Stars accumulate. Forks sit dormant. READMEs go stale. GitX MCP shifts the question to what is happening now — synthesizing commit velocity, pull request throughput, issue engagement, and contributor patterns into a single 0–100 momentum score.

The result is a clear status verdict: active, abandoned, stable, or early alpha. Each verdict comes with a risk level and a set of plain-English warnings that tell you exactly what to watch out for.

GitX MCP exposes all of this through an MCP-compatible HTTP endpoint, so AI coding tools — Cursor, Claude Desktop, Claude Code — can query it directly in context, without the developer leaving their editor.

Example agent queries

These are the kinds of questions an AI agent can answer in context using GitX MCP.

  • 01Which Solana repos are gaining developer momentum this week?
  • 02Analyze this repo and tell me if it is active or abandoned.
  • 03Compare these repositories by developer activity.
  • 04Find early alpha repos in AI agents.
  • 05Is this crypto infrastructure project still maintained?

Momentum signals

The 0–100 momentum score is built from six independent signals.

Commit Velocity

Recent commit counts over 7-day and 30-day windows. A proxy for sustained developer attention rather than burst activity.

Pull Request Activity

Volume of opened and merged PRs in the past 30 days. High PR throughput is the strongest signal of active collaboration.

Issue Engagement

New issues opened and responded to. A maintained repo closes issues; an abandoned one lets them accumulate.

Contributor Breadth

Number of distinct contributors. Single-maintainer repos carry significantly higher bus-factor risk.

Push Recency

Days since last push. A direct indicator of whether any code has moved recently, regardless of PR or issue noise.

Stars & Fork Baseline

Adoption signal. A repo with 5k stars and 200 forks has demonstrated real-world utility — a modest baseline adjustment to the score.

How repo scoring works

Score Breakdown

  • Commits (30d)25 pts
  • Pull Request Activity (30d)20 pts
  • Issue Activity (30d)15 pts
  • Contributor Breadth15 pts
  • Push Recency15 pts
  • Stars + Forks Baseline10 pts

Status Rules

Active

5+ commits, recent push within 30d, score ≥ 40

Stable

Mature (180d+), maintained, score ≥ 15

Early Alpha

Repo < 90 days old, score ≥ 20

Abandoned

No push in 180d+ or extremely low activity

Trending repositories

View all →

The Dashboard tracks developer momentum across Solana, MCP, AI agents, DevTools, and more. Repos are scored in real time against the GitHub API and ranked by momentum, not just stars.

Open Dashboard →

Use with AI coding tools

Add GitX MCP as an MCP server in your AI coding tool. Once connected, your agent can call analyze_repo, find_trending_repos, and compare_repos directly — without leaving the editor.

mcp config (cursor / claude desktop)

mcp.json
{
  "mcpServers": {
    "gitxmcp": {
      "url": "https://your-deployment.vercel.app/api/mcp",
      "env": {
        "GITHUB_TOKEN": "your_github_token"
      }
    }
  }
}

rest api

shell
# Analyze a repository
curl -X POST https://your-deployment.vercel.app/api/analyze \
  -H "Content-Type: application/json" \
  -d '{"owner": "solana-labs", "repo": "solana"}'

# Find trending repos
curl "https://your-deployment.vercel.app/api/trending?topic=mcp&limit=5"
Read the full documentation →