How do I install the Benchmark On Visionanalysis skill?

Run npx skillmds@latest add libreyolo/benchmark-on-visionanalysis in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Benchmark On Visionanalysis safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Benchmark On Visionanalysis?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Benchmark On Visionanalysis free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Benchmark On Visionanalysis?

libreyolo (@libreyolo) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Benchmark On Visionanalysis by libreyolo

Name: Benchmark On Visionanalysis
Author: libreyolo

Benchmark On Visionanalysis

Signpost for benchmarking LibreYOLO models for visionanalysis.org. Use when someone wants to "benchmark for visionanalysis", produce a submission for the site, measure a model on COCO for publication, or add a hardware/runtime row to the site. The actual work lives in two OTHER repos; this skill orients you and hands off. It does not run benchmarks itself.

libreyolo 6262215 3.0 KB Updated Sep 17, 2026

File contents

Benchmark a LibreYOLO model for visionanalysis.org

This is a pointer skill. The benchmark and publish logic live in two separate repos, each with its own authoritative skill. Do not run benchmarks from inside libreyolo; switch to the right repo and follow its skill.

The two repos

Repo	Role	Authoritative skill (open on GitHub)
`vision-analysis-benchmark` (the harness)	Runs models, emits `va.submission.v1` JSON	`generate-benchmark-results`
`vision-analysis` (the website)	Validates JSON, rebuilds the dataset, deploys	`submit-benchmark-results`

Direct links to the authoritative skills (read these — this signpost only orients):

Local checkouts on this machine: C:\Users\Usuario\Documents\GitHub\vision-analysis-benchmark and C:\Users\Usuario\Documents\GitHub\vision-analysis. GitHub: LibreYOLO/vision-analysis-benchmark, LibreYOLO/vision-analysis.

What to know before you start

The dataset, layout gotchas, protocol config, supported runtimes, and the libreyolo_commit rule all live in the harness skill's "Dataset & protocol" section (generate-benchmark-results). Read that there; this signpost does not copy it (so it cannot drift). One thing worth knowing up front: the canonical eval set is the HF dataset LibreYOLO/coco-val2017-mini500, not full COCO.

Reproducibility: each emitted submission (harness >= 2.1.0) carries a repro block recording the exact command, harness + libreyolo commits, a verifiable image-id fingerprint, and weights hash / export manifest. When someone asks how a published number was produced, that block is the answer. See the harness skill's "Reproducing a published result" section for the step-by-step.

Flow

Go to vision-analysis-benchmark, follow generate-benchmark-results. This emits one va.submission.v1 JSON per (model x runtime x hardware) run.
Hand the emitted JSON(s) to vision-analysis, follow submit-benchmark-results to validate, rebuild generated/verified-results.v1.json, and open the PR / deploy.

The two skills above are authoritative. If anything here disagrees with them, they win - update this signpost rather than diverging.

---name: benchmark-on-visionanalysisdescription: Signpost for benchmarking LibreYOLO models for visionanalysis.org. Use when someone wants to "benchmark for visionanalysis", produce a submission for the site, measure a model on COCO for publication, or add a hardware/runtime row to the site. The actual work lives in two OTHER repos; this skill orients you and hands off. It does not run benchmarks itself.---# Benchmark a LibreYOLO model for visionanalysis.orgThis is a **pointer skill**. The benchmark and publish logic live in two separaterepos, each with its own authoritative skill. Do not run benchmarks from inside`libreyolo`; switch to the right repo and follow its skill.## The two repos| Repo | Role | Authoritative skill (open on GitHub) ||---|---|---|| `vision-analysis-benchmark` (the harness) | Runs models, emits `va.submission.v1` JSON | [`generate-benchmark-results`](https://github.com/LibreYOLO/vision-analysis-benchmark/blob/main/skills/generate-benchmark-results/SKILL.md) || `vision-analysis` (the website) | Validates JSON, rebuilds the dataset, deploys | [`submit-benchmark-results`](https://github.com/LibreYOLO/vision-analysis/blob/main/skills/submit-benchmark-results/SKILL.md) |Direct links to the authoritative skills (read these — this signpost only orients):- Harness: https://github.com/LibreYOLO/vision-analysis-benchmark/blob/main/skills/generate-benchmark-results/SKILL.md- Website: https://github.com/LibreYOLO/vision-analysis/blob/main/skills/submit-benchmark-results/SKILL.mdLocal checkouts on this machine:`C:\Users\Usuario\Documents\GitHub\vision-analysis-benchmark` and`C:\Users\Usuario\Documents\GitHub\vision-analysis`.GitHub: `LibreYOLO/vision-analysis-benchmark`, `LibreYOLO/vision-analysis`.## What to know before you startThe dataset, layout gotchas, protocol config, supported runtimes, and the`libreyolo_commit` rule all live in the harness skill's **"Dataset & protocol"**section (`generate-benchmark-results`). Read that there; this signpost does notcopy it (so it cannot drift). One thing worth knowing up front: the canonicaleval set is the HF dataset `LibreYOLO/coco-val2017-mini500`, not full COCO.Reproducibility: each emitted submission (harness >= 2.1.0) carries a `repro`block recording the exact command, harness + libreyolo commits, a verifiableimage-id fingerprint, and weights hash / export manifest. When someone asks howa published number was produced, that block is the answer. See the harnessskill's **"Reproducing a published result"** section for the step-by-step.## Flow1. Go to `vision-analysis-benchmark`, follow `generate-benchmark-results`. This emits one `va.submission.v1` JSON per (model x runtime x hardware) run.2. Hand the emitted JSON(s) to `vision-analysis`, follow `submit-benchmark-results` to validate, rebuild `generated/verified-results.v1.json`, and open the PR / deploy.The two skills above are authoritative. If anything here disagrees with them, theywin - update this signpost rather than diverging.

Frequently asked questions