DeepSpeed is a software library that helps machine-learning systems train and run deep-learning models across multiple GPUs or machines. It is used by researchers and engineers who need distributed training and inference for large models. The catalogue entries provide instructions for using DeepSpeed.
Borrowing it
Nothing to install: this file belongs to deepspeedai/DeepSpeed. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/deepspeedai/DeepSpeed/master/AGENTS.mdgit clone --depth 1 https://github.com/deepspeedai/DeepSpeedWrote 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/deepspeedai/deepspeed/agents-md)<a href="https://agentmods.dev/instructions/deepspeedai/deepspeed/agents-md"><img src="https://agentmods.dev/badge/instructions/deepspeedai/deepspeed/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/deepspeedai/deepspeed/agents-md"><img src="https://agentmods.dev/badge/instructions/deepspeedai/deepspeed/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.01385 | $0.01385 |
| Opus 5.5 | $0.00554 | $0.00554 |
| Sonnet 5.5 | $0.00277 | $0.00277 |
| Haiku 4.5 | $0.00138 | $0.00138 |
Grade A, and why
DeepSpeed 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 12d 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 — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Workspace-level instructions for AI coding agents
These rules apply to all AI-assisted contributions to
deepspeedai/DeepSpeed.
Contribution Policy (Mandatory)
Duplicate-work checks
Before proposing a PR, run these checks:
gh issue view <issue_number> --repo deepspeedai/DeepSpeed --comments
gh pr list --repo deepspeedai/DeepSpeed --state open --search "<issue_number> in:body"
gh pr list --repo deepspeedai/DeepSpeed --state open --search "<short area keywords>"
- If an open PR already addresses the same fix, do not open another.
- If your approach is materially different, explain the difference in the issue.
- DeepSpeed carries long-lived parallel implementations (ZeRO 1/2 vs 3, fp16 vs bf16 optimizers, per-accelerator op builders). Before fixing a bug in one, check whether the sibling paths share the defect, and state in the PR which ones you checked.
New features need an accepted proposal first
Per CONTRIBUTING.md, a new feature starts with a proposal issue -- description, motivation, rough design, and planned experiments -- that maintainers accept before implementation begins. Do not open a feature PR with no accepted proposal; open the proposal issue instead.
No low-value busywork PRs
Do not open one-off PRs for tiny edits (a single typo, an isolated style change, a lone type annotation). Mechanical cleanups are acceptable only when bundled with substantive work in the same area. This is the PR-level form of the rule against cosmetic changes below.
Accountability
- Pure code-agent PRs are not allowed. A human submitter must understand and defend the change end-to-end.
- The submitting human must review every changed line and run the relevant tests.
Signed-off-byis a DCO attestation made by a person, not by an agent. Never sign off on a diff that no human has read.- PR descriptions for AI-assisted work must include:
- Why this does not duplicate an existing PR or issue.
- Test commands run and their results, including the hardware they ran on: device model and count, accelerator/driver version, and torch version.
- Convergence, throughput, and peak-memory results when the change affects numerics, performance, or memory -- see
CONTRIBUTING.mdfor the evidence bar. - Which cells of the compatibility matrix (ZeRO stage, offload, precision, parallelism) were validated, and which were not.
- A clear statement that AI assistance was used.
- Before opening or reviewing a PR, work through the
/pr-checklistskill: design fit, accelerator portability, backward compatibility, on-device validation, diff hygiene, and PR contents.
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.
- 12d ago Changed · +1 lines · +48 tokens per session b348fed14152
- 17d ago Changed · +45 lines · +706 tokens per session 376a7490c155
- 18d ago Changed · -1 lines · -83 tokens per session 00c5df179ec7
- 1mo ago Changed · +1 lines · +57 tokens per session 0308bec9b044
- 1mo ago Changed · +10 lines · +202 tokens per session 33848a172f9e
- 1mo ago First seen · 33 lines · 455 tokens per session scan A 78604afb71eb
DeepSpeed AGENTS.md is an instructions file published in the GitHub repository deepspeedai/DeepSpeed (43,271 stars, last pushed today), licensed Apache-2.0. It adds 1,385 tokens to every session, about $0.0055 per session on Opus 5.5. 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 2026-09-01.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.