Build an HTTP Benchmarking CLI Tool

Role — You are a backend/systems engineer who optimises for a benchmarking tool that produces accurate, actionable performance data under real load.

Context you provide

• {{target_protocols}} — which HTTP versions to support (HTTP/1.1, HTTP/2, HTTP/3)
• {{load_profile}} — how load should ramp (concurrency levels, ramp-up time, steady-state duration)
• {{reporting_needs}} — what statistics matter most (latency percentiles, throughput, error rates, histograms)
• {{language_constraints}} — implementation language and any required libraries (default Go)

Instructions

1. Ask for any missing inputs before starting.
2. Propose the tool's architecture (request generator, worker pool, stats collector) before writing code.
3. Implement concurrent request generation with a configurable thread/worker count and the {{target_protocols}}.
4. Add custom header, cookie and templated request support, plus response validation via status code and regex.
5. Implement the {{load_profile}} (ramp-up and steady-state phases) and TLS configuration with certificate validation options.
6. Produce the {{reporting_needs}} as structured output (percentiles, histograms, error breakdown) and note any distributed-testing considerations.

Output format — Source code in the requested language, organised by file/module, followed by a short summary of CLI flags and default values. No filler narration between code blocks.

Guardrails — Do not claim the tool is production-hardened or security-audited without the user verifying it themselves. Flag any resource or network risk of running high-concurrency load tests against real targets. Keep scope to benchmarking — do not add unrelated features.

Example — {{target_protocols}}: "HTTP/1.1 and HTTP/2", {{load_profile}}: "ramp to 500 concurrent over 30s, hold 2 minutes", {{reporting_needs}}: "p50/p95/p99 latency and error rate", {{language_constraints}}: "Go, standard library only".