model-benchmarking

model-benchmarking is a skill for Claude Code from RBraga01/builder-ai. It costs 43 tokens per session (1,254 once invoked), scanned A, original, MIT.

A model-testing process for choosing which language model should run a production feature. It compares models on your own task examples using defined measures.

In plain words
What is it for?
Use it to compare providers, assess cheaper or faster replacements, and decide which model performs well on a specific production task.
Why use it?
It prevents choosing a model from general leaderboard scores that may not reflect your application's needs.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: positional $N argument.

Part of the builder-ai plugin — 8 skills, 5 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/rbraga01/builder-ai/model-benchmarking
Any agent
npx skills add RBraga01/builder-ai --skill model-benchmarking
Clone the repo
git clone --depth 1 https://github.com/RBraga01/builder-ai

Made for: Claude Code.

Or install builder-ai, the plugin that ships this one along with the rest of its 8 skills, 5 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for model-benchmarking

README.md
[![agentmods](https://agentmods.dev/badge/skills/rbraga01/builder-ai/model-benchmarking.svg)](https://agentmods.dev/skills/rbraga01/builder-ai/model-benchmarking)
Your own site
<a href="https://agentmods.dev/skills/rbraga01/builder-ai/model-benchmarking"><img src="https://agentmods.dev/badge/skills/rbraga01/builder-ai/model-benchmarking.svg" alt="Measured on agentmods" height="20"></a>
Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,254 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00043 $0.01254
Opus 5 $0.00022 $0.00627
Sonnet 5 $0.00009 $0.00251
Haiku 4.5 $0.00004 $0.00125

Measured 5d ago against content hash 7cd95fd00299, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

model-benchmarking scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/model-benchmarking/SKILL.md · 121 lines

How it starts

The opening of the file, as written. The whole thing — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Model Benchmarking

The Law

A MODEL IS NOT CHOSEN UNTIL IT HAS BEEN TESTED ON YOUR TASK DATA.
Leaderboard scores are averaged over tasks you're not building.
"It's the best model" is a claim about benchmarks someone else ran.
Task data + defined metric + three tested tiers IS a model selection.

When to Use

Trigger when:

  • Choosing a model for any new production feature
  • Considering a switch to a newer, cheaper, or faster model
  • Validating that a smaller model can replace a larger one
  • Comparing providers (OpenAI, Anthropic, Mistral, local)

When NOT to Use

  • Initial feasibility exploration (before task definition is stable) — benchmark when you know what you're measuring
  • Model compatibility checks (does this model support tool use, JSON mode, etc.) — that's a capability query, not a benchmark
  • Leaderboard research to narrow the candidate list — that's input to Stage 3, not a substitute for it

The Process

Four stages. Do not collapse them.

Stage 1 — Define the Bar

Before running any model, write down:

Decision What to Define
Task Exact input format, exact output format, edge cases
Metric Accuracy, faithfulness score, extraction F1, LLM-as-judge, task pass rate
Pass threshold The minimum score to go to production (e.g., ≥ 88%)
Budget constraint Max cost per 1k calls, max monthly spend
Latency constraint Max acceptable p95 (e.g., ≤ 2.0s)

If you cannot define the metric and threshold first, the selection criteria are not clear enough to proceed.

Stage 2 — Build a Representative Test Set

  • Minimum 50 examples; 200+ for features used at volume
  • Label ground truth before running any model. Seeing outputs first contaminates labels.
  • Include: easy cases, hard cases, edge cases, adversarial inputs
  • Tag each example by difficulty: easy / medium / hard
  • Store in benchmarks/<feature>/test-set.jsonl

Stage 3 — Benchmark at Least Three Tiers

Always test at least one tier below your expected choice:

Read the full file on GitHub · 121 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 121 lines · 43 tokens per session scan A 7cd95fd00299

Subscribe to this mod's changes

model-benchmarking is a skill published in the GitHub repository RBraga01/builder-ai (2 stars, last pushed 1mo ago), licensed MIT. It adds 43 tokens to every session and 1,254 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

ai-engineering-toolkit

6 production-ready AI engineering workflows: prompt evaluation (8-dimension scoring), context budget planning, RAG pipeline design, agent security audit (65-point checklist), eval harness building, and product sense coaching.

sickn33/agentic-awesome-skills · 47 tokens

promptfoo-evals

Write, refine, run, and QA promptfoo evaluation suites: promptfooconfig.yaml, prompts, providers, vars, tests, assertions, model-graded rubrics, transforms, datasets, exports, and CI gates. Use for non-redteam eval coverage, regression tests, or new eval matrices. Do not use for adversarial redteam plugin or strategy…

promptfoo/promptfoo · 79 tokens

prompt-master

Generates optimized prompts for AI tools. Activates only when the user explicitly asks to write, fix, improve, or adapt a prompt for a specific AI tool (LLM, Cursor, Midjourney, image AI, video AI, coding agents, etc.). Does not activate for general conversation, coding tasks, document writing, or other…

nidhinjs/prompt-master · 78 tokens

flux2-lora-training

Comprehensive reference for training LoRAs on FLUX.2 Klein 9B and Qwen Image Edit 2511 models. Use this skill whenever the user asks about: training LoRAs for flux2/flux 2 klein/qwen-image-edit, before/after edit LoRAs (head swap, face swap, image editing), inpainting LoRAs, training at larger resolutions, latent…

AnastasiyaW/codex-claude-code-config · 242 tokens

deepseek-provider-contract

Validate a proposed DeepSeek API integration before any key or project context is sent: check thinking-mode tool-call history, strict-schema assumptions, bounded output, and provider data boundaries. Use when integrating DeepSeek, adding DeepSeek tool calls or streaming, debugging DeepSeek 400 after a tool call, or…

AnastasiyaW/codex-claude-code-config · 0 tokens

forensic-prompt-compiler

Forensic image-to-prompt compiler for image generation models. Use this skill whenever the user wants to: convert/describe an existing image into a generation prompt, reconstruct a scene as a prompt, generate prompts from reference images for AI image tools (Midjourney, FLUX, Stable Diffusion, DALL-E, or any diffusion…

AnastasiyaW/codex-claude-code-config · 218 tokens