Create perf-only autoresearch scaffolds with autoresearch.toml, program.md, and eval.py.
npx skills add https://github.com/Elijah-J/InfoScraper --skill ar-new
Please help me install this Agent Skill. Skill: ar-new Source: https://github.com/Elijah-J/InfoScraper/tree/main/.agents/skills/ar-new Command: npx skills add https://github.com/Elijah-J/InfoScraper --skill ar-new
Create a perf-only autoresearch scaffold to jumpstart structured benchmark experiments.
Branch autoresearch/$ARGUMENTS and initialize the scaffold for a perf-only autoresearch project.
High-intent search queries and answers about installing and using this skill.
To scaffold a performance benchmarking project, you generate an autoresearch.toml skeleton alongside program.md and eval.py files to standardize optimization loops and enforce a clear evaluation contract with defined metrics and guardrails.
An autoresearch scaffold is a standardized project structure for early-stage optimization experiments. You need one when starting perf-focused benchmarking loops that require clear metrics, guardrails, and mutable surfaces to ensure repeatable workflows and structured results.
You standardize performance experiments by generating essential files like an autoresearch.toml skeleton, program.md, and eval.py within a dedicated branch, enforcing a repeatable workflow with a defined results structure and clear evaluation contract across teams.
No, this scaffold applies specifically to early-stage optimization experiments that already have clear metrics, guardrails, and mutable surfaces defined. It requires these elements to enforce its repeatable workflow and structured evaluation contract.
To start a perf-only autoresearch project, you need an autoresearch.toml skeleton plus program.md and eval.py files. These establish the defined results structure and evaluation contract required for the benchmarking loop.
The best way to initialize a perf-focused autoresearch project is to branch into autoresearch/$ARGUMENTS and trigger the scaffold generation, which creates the necessary autoresearch.toml skeleton and evaluation files for rapid iteration.
yu13130122297Identify uncertainties in research framing, datasets, metrics, and baselines.
JamesPrialCreate Go benchmarks with b.N loops, sub-benchmarks, and timing control.
gofhirProfile Go applications with benchmarks and CPU profiling to reduce allocations and latency.
kousukefAutomate creation, modification, evaluation, and benchmarking of Claude skills.
