Comments
>log in to commentI built hotpath-rs to make it easier for Rust developers to understand where their applications spend time and resources. The goal is simple: one easy-to-use tool to identify different performance bottlenecks - CPU, memory, async, SQL, HTTP, I/O, locks, and channels - without juggling multiple profilers. I'm also working on GitHub CI integration to catch performance regressions before they reach production. Would love to hear your feedback and what performance challenges you're facing in your Rust projects! 🦀
- Profiles CPU, memory, async, SQL, HTTP, I/O and lock contention
- Exports metrics to Prometheus and Grafana
- TUI demo available via SSH (demo.hotpath.rs)
- AI agents can auto‑configure profiling (e.g., Claude, Codex)
- Supports live monitoring and static reports
hotpath-rs is an open-source Rust profiler that tracks CPU, memory, async, I/O, SQL and HTTP performance with live TUI and Prometheus/Grafana export.
- for
- Rust developers and AI coding agents who need detailed performance insights.
- pricing
- open source
- license
- MIT
Key features
- Time, CPU & memory profiling — Measures function execution time, allocation hotspots and detects memory leaks.
- Async observability — Tracks futures, channels, streams and Tokio runtime metrics.
- I/O and network monitoring — Reports bytes, throughput and latency for files, TCP, Redis and compression streams.
- SQL & HTTP profiling — Provides per-query latency for sqlx/Diesel and per-endpoint metrics for reqwest/ureq.
- Prometheus & Grafana integration — Exports all profiling metrics for production dashboards.
- Live TUI dashboard — Shows real-time performance data and async data flow in a terminal UI.
- CI regression detection — Runs automated benchmarks on every pull request via GitHub CI.
- MCP server for AI agents — Allows LLMs to query profiling data in real time via Model Context Protocol.
Use cases
- Identify CPU-bound vs I/O-bound functions in a Rust service
- Detect performance regressions on each PR in CI pipelines
- Monitor production latency and throughput of HTTP endpoints with Grafana
hotpath-rs FAQ
How do I add hotpath to an existing Rust project?+
Install the CLI with `cargo install hotpath`, add the `hotpath` dependency to Cargo.toml, enable the `hotpath` feature, and annotate functions with `#[hotpath::measure]`.
Can hotpath run without adding compile-time overhead?+
Yes, profiling is feature-gated; when the `hotpath` feature is disabled the macros become no-ops and add no overhead.
How does hotpath integrate with CI?+
Use the built-in CI integration to benchmark each pull request automatically, detecting regressions without extra configuration.
What visualization options are available?+
You can generate static reports after program exit or use the live TUI dashboard for real-time metrics; both can export to Prometheus for Grafana dashboards.
Is hotpath usable by AI agents?+
Yes, the `hotpath init --agent <name>` command lets Claude, Codex or OpenCode configure profiling automatically in a repository.
Which async runtimes are supported?+
Hotpath provides Tokio runtime monitoring, including worker stats and task scheduling.
Summarized by DevHunt from hotpath.rs · Oct 20, 2026. Details may change; check the official site.
