skills.rest
lespaceman
lespacemanCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

run-benchmark

Automate WebBench benchmark execution and generate results in JSONL format.

4|1|Updated Feb 3, 2026
Search Tags:#automation#testing#reporting#benchmark#browser#web-bench
One-click install
npx skills add https://github.com/lespaceman/athena-workflow-marketplace --skill run-benchmark-lespaceman
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: run-benchmark
Source: https://github.com/lespaceman/athena-workflow-marketplace/tree/main/plugins/web-bench/skills/run-benchmark
Command: npx skills add https://github.com/lespaceman/athena-workflow-marketplace --skill run-benchmark-lespaceman

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates end-to-end browser benchmark orchestration, reducing manual setup and coordination overhead for performance testing.

Core Features & Use Cases

  • Orchestrates dataset loading, task execution, evaluation, and reporting in a single run.
  • Supports configurable category filters, sample sizes, resume from previous runs, and report-only modes.
  • Generates structured results and summaries to speed up benchmarking workflows.

Quick Start

Run the benchmark using the primary skill invocation with appropriate flags to select category and sample size.

Frequently Asked Questions about run-benchmark

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate end-to-end browser benchmark execution and reporting?▼

Automate end-to-end browser benchmark execution by orchestrating dataset loading, task execution, evaluation, and report generation within a single interactive run. It configures category filters and sample sizes, then writes structured results to web-bench-results.jsonl.

Can I resume a web benchmark from a previous run without starting over?▼

Yes, you can resume a web benchmark from a previous run by passing the --resume command-line argument. This allows you to continue task execution and evaluation without losing prior progress or reloading completed dataset samples.

How do I filter browser benchmark tasks by category and sample size?▼

Filter browser benchmark tasks by category and sample size using the --category and --sample command-line arguments. This configures the dataset loading and task execution modules to run only the specified subset of web bench tests.

Is it possible to generate benchmark reports without executing new tasks?▼

Yes, generate benchmark reports without executing new tasks by using the --report-only mode. This skips dataset loading and task execution, directing the tool to solely generate a summary from existing evaluation results.

Does this browser benchmarking tool require external dependencies?▼

No, this browser benchmarking tool requires no external dependencies to run. It coordinates internal modules for loading datasets, executing tasks, evaluating results, and generating reports directly within a single interactive session.

Related Skills

Explore More Software Engineering Skills→
ResearAIResearAI
3.3k

baseline

Standardize a reference object for quest evaluation with plan and checklist requirements.

Community
Advanced
lespacemanlespaceman
4

generate-report

Aggregate WebBench results from JSONL into a markdown report.

Community
Advanced
exceptionlessexceptionless
2.5k

agent-browser

Automate browser interactions via command-line interface for web tasks.

Official
Advanced
binfen1binfen1

benchmark

Measure performance regressions with browse daemon baselines across PRs.

Community
Advanced