Endpoint Latency Profiler
Measure API response times, compare with benchmarks, and identify bottlenecks
Example Prompt
Profile the latency of all our public API endpoints and flag anything slower than 500ms
About
Profiles API endpoint performance by fetching each endpoint and measuring response times, querying historical latency data from your database for comparison, and identifying endpoints that exceed benchmark thresholds. The profiler produces a detailed report with p50/p95/p99 latency comparisons and bottleneck analysis.
Workflow Steps
Fetch each API endpoint to measure current response times and status codes
🌐 Fetch URLQuery historical latency percentiles from the metrics database for baseline comparison
📊 Run SQL QueryReason through the latency data to identify bottlenecks and degradation patterns
🧠 Structured ReasoningWrite a latency profiling report with benchmarks, current measurements, and recommendations
✏️ Write FileTools Used
Fetch URL
Make an HTTP request to a URL and return the response body
Write File
Write or overwrite a file at a specified path with the given content
Run SQL Query
Execute a read-only SQL query against a PostgreSQL database and return results
Structured Reasoning
Break down a complex problem into sequential reasoning steps with explicit chain-of-thought
Required MCP Servers
Fetch
Retrieve and convert web content to markdown, extract text from URLs and APIs
Filesystem
Read, search, and manage files on the local filesystem
PostgreSQL
Query, inspect, and manage PostgreSQL databases with full SQL support
Sequential Thinking
Dynamic problem-solving through structured thought sequences with branching and revision