kstep AGENTS.md

A guide for reproducing Linux kernel scheduling bugs from a specific commit, then developing and testing a matching reproduction.

In plain words
What is it for?
Use it to inspect a kernel patch, trace the affected scheduler code, retrieve linked mailing-list discussions, and build a reproduction driver.
Why use it?
It gives agents a structured way to identify the exact conditions that trigger a kernel bug before attempting a fix or reproduction.

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/kstep-dev/kstep/agents-md
Clone the repo
git clone --depth 1 https://github.com/kstep-dev/kstep

Made for: Codex, OpenCode.

Per session 796 This file is loaded in full into every session.
When invoked 796 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.00796 $0.00796
Opus 5 $0.00398 $0.00398
Sonnet 5 $0.00159 $0.00159
Haiku 4.5 $0.00080 $0.00080

Measured yesterday against content hash 1ca350defdbb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

kstep AGENTS.md scanned grade A with 1 finding 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 yesterday.

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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

`curl -sL https://lore.kernel.org/lkml/<email_id>/t.mbox.gz | gunzip`
AGENTS.md · 45 lines

How it starts

The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Bug Reproduction

To reproduce a bug fixed in commit [hash], follow these steps:

Planning Stage
  • Check out the Linux source code just before the fix: ./checkout.py [hash]~1 [name]_buggy --git.
  • Read the commit message and patch carefully: git -C build/master show -U32 [hash].
  • Carefully analyze the changes introduced in the patch, focusing on modifications in build/master/kernel/sched. Trace the call chain that leads to the bug, ensuring you understand the purpose of each function and how they interact. Clearly identify the conditions and system state required to trigger the bug.
  • If the commit message contains links, carefully review the content of ALL links as they may provide the additional information about the bug (especially the reproduction conditions). For LKML links (e.g., lore.kernel.org, lkml.kernel.org, patch.msgid.link), extract the email ID and retrieve the message with:
    curl -sL https://lore.kernel.org/lkml/<email_id>/t.mbox.gz | gunzip
  • Thoughtfully analyze the conditions and mechanisms that trigger the bug, and plan an effective method to reproduce it within the kernel.
Development Stage
  • Create a driver in kmod/drivers_generated/[driver_name].c that triggers the bug.
  • Do not leave the driver broken. There could be other agents running with you.
  • Guard the driver with #if LINUX_VERSION_CODE to the compatible kernel version. if unsure, just to the current version you are using.
  • Try not to directly manipulate internal scheduler state.
  • Add detailed logging in your driver for all relevant fields to aid debugging. If necessary, add kernel-side logging with printk().
  • Build and execute the driver on the buggy kernel with: ./checkout.py [hash]~1 [name]_buggy --git && make linux NAME=[name]_buggy && ./run.py [driver_name] --name [name]_buggy
  • Determine whether the bug is reproduced by examining the output logs: cat data/logs/latest.log
  • If you do not observe the bug, refine your implementation and repeat the process.
  • After confirming that your driver triggers the bug, rerun the same driver on the fixed kernel: ./checkout.py [hash] [name]_fixed --git && make linux NAME=[name]_fixed && ./run.py [driver_name] --name [name]_fixed
  • Review the logs to ensure the bug no longer occurs. In rare cases where the issue persists despite the fix, provide clear documentation and report your findings.
Refinement Stage
  • If you manually altered internal scheduler state to trigger the bug during the triggering stage, refine your driver to reproduce the bug using only public kernel APIs or kSTEP interfaces (kmod/driver.h). If needed, consider extending kSTEP to provide the required functionality.
  • Make sure that the driver is deterministic, and produce the same trace on the same kernel. If not, investigate the root cause of the non-determinism.
  • After confirming your driver can trigger the bug, clearly demonstrate its impact through externally observable behavior, such as changes in task scheduling, rather than relying solely on kernel-internal state.
  • Once complete, add your driver to BUGS_EXTRA in reproduce.py to enable automated testing.

Read the full file on GitHub · 45 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. yesterday First seen · 45 lines · 796 tokens per session scan A 1ca350defdbb

Subscribe to this mod's changes

kstep AGENTS.md is an instructions file published in the GitHub repository kstep-dev/kstep (11 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 796 tokens to every session, about $0.0040 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other instructions, from other repositories

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

buildNext

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).

microsoft/vscode · 6,785 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

spec-kit AGENTS.md

Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,040 tokens

langchain 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.

langchain-ai/langchain · 4,345 tokens