Automate Capsem performance benchmarking in a VM and output JSON reports.
npx skills add https://github.com/google/capsem --skill dev-benchmark
Please help me install this Agent Skill. Skill: dev-benchmark Source: https://github.com/google/capsem/tree/main/skills/dev-benchmark Command: npx skills add https://github.com/google/capsem --skill dev-benchmark
Capsem benchmarking helps developers measure performance characteristics of Capsem components inside isolated VMs, enabling detection of regressions and validation of optimizations before release.
Run the full benchmark suite inside the VM by executing just bench and inspect the JSON report at /tmp/capsem-benchmark.json.
High-intent search queries and answers about installing and using this skill.
VM performance benchmarking across disk, startup, and http is automated by running the full benchmark suite inside the VM. It executes all seven categories and outputs structured timing and throughput metrics to a JSON report.
Detecting Capsem performance regressions involves running automated benchmarks inside the VM and comparing the JSON output over time. This validates optimizations and identifies slow operations before release.
Yes, capsem-bench must be available inside the VM to execute performance benchmarks. The tool runs in-VM and saves detailed timing and throughput results to /tmp/capsem-benchmark.json.
Yes, individual benchmark categories including disk, rootfs, startup, http, throughput, and snapshot can be run independently. You can also select 'all' to execute the complete benchmark suite.
In-VM benchmarking outputs structured metrics by generating a JSON report at /tmp/capsem-benchmark.json. This file details timing and throughput data suitable for automated tooling and performance dashboards.
Slow operations after recent changes can be investigated by running benchmarks to compare current performance against previous JSON reports. This helps isolate whether new changes introduced performance regressions.
googleRun capsem-doctor diagnostics and manage test fixtures inside the guest VM.
olaafrossiBenchmark web pages against baselines and report performance regressions.
ougi777Diagnose local host resource bottlenecks by analyzing CPU, memory, and disk usage.
ryukyagamilightProfile CPU and memory usage with perf, valgrind, and strace.
