all scenarios

Flaky (50% errors)

flaky-500 · Status & reliability

Endpoint (Streamable HTTP)
https://chaos-mcp.thomascdnns.com/s/flaky-500/mcp
On the wire — recorded from a real request
your clientchaos-mcp01initializePOST /s/flaky-500/mcp200 OK · application/json · 1 ms"jsonrpc""2.0""id"1"result""protocolVersion""2024-11-05""capabilities""tools""serverInfo""name""chaos-mcp/flaky-500""version""1.0.0""instructions""chaos-mcp scenario: flaky-500 — Flaky (50% errors)"02tools/listPOST /s/flaky-500/mcp200 OK · application/json · 1 ms"jsonrpc""2.0""id"2"result""tools""name""echo""description""Echo back the provided message.""inputSchema""type""object""properties""message""type""string""description""Text to echo"… 40 more lines

Symptom

what you'll observe

Roughly half of requests fail with 500; retries sometimes succeed.

Cause

what this reproduces

One bad replica behind a load balancer, or a race that trips intermittently.

Fix

how you'd resolve it

Make handlers deterministic; add readiness probes so bad replicas leave rotation.

Try it with curl
bash
curl -sS -X POST \
  https://chaos-mcp.thomascdnns.com/s/flaky-500/mcp \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json' \
  -d '
"jsonrpc""2.0"
"id"1
"method""initialize"
"params"
"protocolVersion""2024-11-05"
"capabilities"
"clientInfo""name""curl""version""1"

'