New Version 1.2.0 is out

Introduction to APIsage

APIsage is an intelligent API analysis platform designed for the frontier. Our MCP (Model Context Protocol) server enables LLMs to interact directly with your infrastructure for real-time diagnostics and endpoint validation.

download_for_offlineInstallation

To get started, visit mcpize.com/mcp/apisage to obtain your API Key. Once you have your key, follow the configuration steps for your preferred client below.

CLI Command for Claude Desktop:

claude mcp add --transport http apisage https://apisage.mcpize.run --header "API_KEY: ••••••••••••••••••••••••••••••••"

Or paste this JSON config ▾

{
  "mcpServers": {
    "apisage": {
      "type": "http",
      "url": "https://apisage.mcpize.run",
      "headers": {
        "API_KEY": "your_api_key_here"
      }
    }
  }
}

Config file: Claude Code: .mcp.json (project) or ~/.claude.json (user). Claude Desktop needs an mcp-remote bridge for HTTP servers.

terminalUsage Examples

Discover how to leverage the core tools of APIsage through practical, real-world scenarios.

01. Quick Endpoint Health Check

Input
"Check if the /v1/users endpoint is responding correctly."
Output (JSON)
{
  "status": "healthy",
  "endpoint": "/v1/users",
  "latency": "142ms",
  "response_code": 200,
  "payload_valid": true
}

02. Diagnose a Failing Endpoint

Input
"Why is the /checkout endpoint returning a 500 error?"
Output (Analysis)

Critical: Internal Server Error detected.

Trace analysis reveals a NullPointerException in the PaymentGatewayService.java:142. The upstream credit service timed out, and the error handler failed to catch the specific exception.

Recommended Fix:
  • Implement circuit breaker for upstream dependencies.
  • Add null-check for gateway response object.

03. Detect Backend Framework

Input
"Analyze the headers of api.example.com to identify the tech stack."
Output (JSON)
{
  "server": "Node.js/18.x",
  "framework": "Fastify",
  "cache_strategy": "Redis (Cloudflare Edge)",
  "language": "TypeScript"
}

04. Security Audit

Input
"Run a quick security audit on the production auth flow."
Output (Markdown)
warningHigh: JWT Token missing 'HttpOnly' flag.
infoMedium: Missing 'Strict-Transport-Security' header.
check_circleLow: Rate limiting active (100req/min).

05. Performance Analysis

Input
"What is the p99 latency for our Search API under 100 req/s load?"
Output (Stats)
Mean Latency
42ms
p99 Latency
1,240ms
Identified bottleneck: Database indexing on the 'query' parameter is inefficient.

06. Generate API Documentation

Input
"Generate a Swagger/OpenAPI spec for the existing /analytics endpoint."
Output (YAML)
openapi: 3.0.0
info:
  title: Analytics API
  version: 1.0.0
paths:
  /analytics:
    get:
      summary: Retrieve traffic metrics
      parameters:
        - name: range
          in: query
          schema:
            type: string
            enum: [24h, 7d, 30d]

07. Generate Test Cases

Input
"Create Jest tests for the /login endpoint, including edge cases."
Output (Code)
test('returns 401 on invalid password', async () => {
  const res = await request(app)
    .post('/login')
    .send({ email: 'test@user.com', pass: 'wrong' });
  expect(res.statusCode).toEqual(401);
});

test('handles rate limiting after 5 attempts', ...);

08. Compare Endpoints (v1 vs v2)

Input
"What changed between /v1/search and /v2/search?"
Output (Diff)
+Added support for fuzzy matching parameter.
-Deprecated 'legacy_id' field in response.
~Improved response time by 40% using pre-caching.

09. Full Analysis (Orchestrator)

Input
"Run a full infrastructure analysis and report any anomalies."
Output (Report Summary)
Orchestrator finished in 8.4s. Scanned 12 clusters.
Anomaly Found:

Cluster US-EAST-1 shows a 15% increase in CPU usage correlated with the latest deployment of the 'image-processor' microservice. Leak suspected in memory-mapped buffer.

Data Server

Real-time Analysis

APIsage monitors your production environment with millisecond precision, catching anomalies before they impact users.