relarena AGENTS.md

relarena AGENTS.md is an instructions file for Codex, OpenCode from PriorLabs/relarena. It costs 1,258 tokens per session, scanned A, original, Apache-2.0.

Development instructions for RelArena, a package for adding, tuning, and evaluating machine-learning models. They explain how model folders, public interfaces, shared code, optional dependencies, tests, and CPU-cache preparation should be organized.

In plain words
What is it for?
Adding or changing models, choosing where shared code belongs, using the public API, preparing expensive preprocessing caches, and running the required macOS checks.
Why use it?
They prevent models and benchmark-only code from becoming tightly coupled, and help contributors follow stable public interfaces when changing the package.

Instructions file for CodexOpenCode

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 instructions/priorlabs/relarena/agents-md
Clone the repo
git clone --depth 1 https://github.com/PriorLabs/relarena

Made for: Codex, OpenCode.

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 relarena AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/priorlabs/relarena/agents-md.svg)](https://agentmods.dev/instructions/priorlabs/relarena/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/priorlabs/relarena/agents-md"><img src="https://agentmods.dev/badge/instructions/priorlabs/relarena/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,258 This file is loaded in full into every session.
When invoked 1,258 The same file — it is already loaded in full.
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 $0.01258 $0.01258
Opus 5 $0.00629 $0.00629
Sonnet 5 $0.00252 $0.00252
Haiku 4.5 $0.00126 $0.00126

Measured today against content hash f3ba91c837d0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

relarena AGENTS.md 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 today.

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.

AGENTS.md · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

relarena — agent notes

Start with the README for what the package is and how a run works.

This repo is public

Everything committed here is published under Apache-2.0 — code, docs, and baseline_results/ alike. Nothing may reference Prior Labs–internal infrastructure, systems, names, or paths. This applies to everything pushed to the remote, not just tracked files: commit messages, branch names, and PR text included.

Adding or changing a model

Read docs/adding-a-model.md first. Model development is two things: the model folder (layout, fit/predict contract and its datatypes, search-space choices, shared-code placement, optional dependencies, vendoring, tests) and — for methods with expensive CPU pre-processing — a public cache-warm script. models/lightgbm/ is the smallest complete example to copy.

Refactor-friendly structure rules

Keep the contract/tuning core, models, predictive interface, and benchmark evaluation layers cleanly separated:

  • Import through the public API: external-facing names are re-exported in relarena/__init__.py (or a subpackage __init__.py), and docs, examples, and tests use those paths. Underscore-prefixed modules are internal and may move without notice.
  • Respect the layering: model packages and userdb/ import the generic core (model, registry, search_space, tuner, split types) and the shared infrastructure (featurization/) — never each other, and never the benchmark-only evaluation/ subpackage. models/_shared/ is for code shared between models; nothing outside models/ imports it.
  • Keep generic and benchmark-specific code in separate modules: anything touching relbench.datasets/relbench.tasks or data checksums is benchmark code and doesn't belong in model/tuner/search_space.
  • Heavy optional deps are lazy-imported inside fit behind extras; no import-time side effects beyond explicit register_model calls.

macOS: run tests / CLI with OMP_NUM_THREADS=1

Read the full file on GitHub · 94 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. today Changed · +8 lines · +76 tokens per session f3ba91c837d0
  2. 5d ago First seen · 86 lines · 1,182 tokens per session scan A 4453c6a82f20

Subscribe to this mod's changes

relarena AGENTS.md is an instructions file published in the GitHub repository PriorLabs/relarena (57 stars, last pushed yesterday), licensed Apache-2.0. It adds 1,258 tokens to every session, about $0.0063 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 2026-08-30.