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.
npx agentmods add instructions/amd-agi/agentkernelarena/agents-mdgit clone --depth 1 https://github.com/AMD-AGI/AgentKernelArenaWrote 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.
[](https://agentmods.dev/instructions/amd-agi/agentkernelarena/agents-md)<a href="https://agentmods.dev/instructions/amd-agi/agentkernelarena/agents-md"><img src="https://agentmods.dev/badge/instructions/amd-agi/agentkernelarena/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01754 | $0.01754 |
| Opus 5 | $0.00877 | $0.00877 |
| Sonnet 5 | $0.00351 | $0.00351 |
| Haiku 4.5 | $0.00175 | $0.00175 |
Grade A, and why
AgentKernelArena 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Purpose
AgentKernelArena is a controlled environment for evaluating GPU-kernel agents. Every change must preserve reproducibility, benchmark integrity, meaningful correctness checks, and fair baseline-versus-candidate comparison.
Read README.md and CONTRIBUTING.md before making broad changes. Use repository-relative paths in code, configuration, docs, and instructions; never embed paths from a local checkout.
Repository map and sources of truth
main.py: run orchestration, resume, and parallel scheduling.src/: shared prompting, workspace setup, evaluation, scoring, and reporting.agents/: isolated agent integrations and their own configuration.tasks/: self-contained kernel tasks and task-local runners.example_configs/: run-level agent, task, and GPU selections.src/tools/perf/: canonical benchmark helpers materialized into workspaces.src/eval_tools/: evaluator-side optional analysis-tool control plane.tests/: framework, integration, and regression tests.
When behavior and prose disagree, inspect the implementation. In particular:
- Agent identifiers and loading:
src/module_registration.py - Task contract: the task's
config.yaml - Evaluation and scoring:
src/evaluator.py,src/performance.py,src/score.py - Harness protection:
src/harness_guard.py - Benchmark helpers:
src/tools/perf/ - Evaluation-tool policy:
docs/how-to/use-evaluation-tools.md
Update relevant docs in the same change when behavior or public configuration changes. Avoid copying volatile model IDs, image tags, or complete registries into new guidance; link to their source of truth instead.
Architecture boundaries
Keep agents and tasks independent.
- A task defines a backend-neutral contract through its files and
config.yaml. Task code must not import fromagents/or depend on a particular agent, model, prompt format, provider, or authentication mechanism. - Agent implementations must not import task modules, mutate committed task sources, or special-case task names and paths. Add general capability through task configuration and shared framework interfaces instead.
- Keep provider- and CLI-specific behavior under
agents/<agent_name>/. Move code intosrc/only when it is genuinely shared by multiple integrations. - Isolated tasks must remain self-contained after being copied into a workspace;
they must not rely on imports from this repository's
src/tree.
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.
- 2d ago First seen · 184 lines · 1,754 tokens per session scan A 76d5a80a55e7
AgentKernelArena AGENTS.md is an instructions file published in the GitHub repository AMD-AGI/AgentKernelArena (112 stars, last pushed today), licensed Apache-2.0. It adds 1,754 tokens to every session, about $0.0088 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 2026-09-03.
Other instructions, from other repositories
any-agent AGENTS.md
AGENTS.md instructions for mozilla-ai/any-agent, covering claude.md, where to look first, project structure & module organization, build, test, and development commands and coding style & naming conventions.
eval-view AGENTS.md
AGENTS.md instructions for hidai25/eval-view, covering evalview agent instructions, what evalview is, core concepts, testcase and evaluationresult.
godmode AGENTS.md
Instructions for thiientv/godmode, covering godmode contributor instructions, working rules and verification.
cloudbox AGENTS.md
Instructions for acoyfellow/cloudbox, covering cloudbox agent guide, canonical commands, deploy model, cloudflare account boundaries and required production secrets.
hermes-skilleval AGENTS.md
AGENTS.md instructions for Raidriar7170/hermes-skilleval, a project described as: Verification-gated skill routing and self-improvement harness for Hermes-style agent skills.
unitares AGENTS.md
AGENTS.md instructions for cirwel/unitares, covering agents.md — unitares, codex-specific wiring, slash commands (commands/.md), local continuity cache and watcher visibility is manual.