dos-kernel GEMINI.md

dos-kernel GEMINI.md is an instructions file for Gemini CLI from anthony-chaudhary/dos-kernel. It costs 606 tokens per session, scanned A, original, MIT.

Instructions for DOS, a trust-checking system for coding agents that verifies work using observable facts such as Git history, files, and time. It tells agents to check claims instead of relying on their own reports.

In plain words
What is it for?
Use it to verify plans or phases, audit whether a commit matches its changes, and check whether agents’ file edits overlap.
Why use it?
It helps detect unfinished work, misleading commit messages, and conflicting edits before they are reported or allowed to proceed.

Instructions file for Gemini CLI

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/anthony-chaudhary/dos-kernel/gemini-md
Clone the repo
git clone --depth 1 https://github.com/anthony-chaudhary/dos-kernel

Made for: Gemini CLI.

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 dos-kernel GEMINI.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/anthony-chaudhary/dos-kernel/gemini-md.svg)](https://agentmods.dev/instructions/anthony-chaudhary/dos-kernel/gemini-md)
Your own site
<a href="https://agentmods.dev/instructions/anthony-chaudhary/dos-kernel/gemini-md"><img src="https://agentmods.dev/badge/instructions/anthony-chaudhary/dos-kernel/gemini-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 606 This file is loaded in full into every session.
When invoked 606 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 $0.00606 $0.00606
Opus 5 $0.00303 $0.00303
Sonnet 5 $0.00121 $0.00121
Haiku 4.5 $0.00061 $0.00061

Measured 4d ago against content hash 944f9f207014, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dos-kernel GEMINI.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 4d 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.

GEMINI.md · 57 lines

How it starts

The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.

DOS — the part that doesn't believe the agents

This extension gives you the DOS trust syscalls as tools. DOS exists because an agent's word about its own work is the weakest signal in the loop. A worker can say it shipped a fix, say the tests pass, say it booked the reservation — and be wrong, or lying, with full confidence. DOS answers those questions from ground truth instead: git history, the file tree, a clock. Never from what the agent said it did.

Use these tools before you trust a claim, not after.

The rule

When you — or another agent, or the user — claim that a piece of work is done, do not trust the claim. Check it. The tools below check it for you.

  • Before you report a task, phase, or feature as finished, call dos_verify (did this plan/phase actually ship?) or dos_commit_audit (does this commit's message match what its diff actually did?). If the witness does not confirm the claim, say so plainly — do not report done.
  • Before two agents edit the same area at once, call dos_arbitrate to find out whether their file trees collide. Disjoint trees may run together; an overlap is a refuse, and the tool names a free lane instead.
  • When you must decline or report blocked, use dos_refuse_reasons / dos_check_reason to pick a structured reason from the closed set, instead of free-text prose a system can't verify.
  • Before you rely on a cited legal case, call dos_citation_resolve to confirm it exists in a real reporter — the antidote to a fabricated citation.

Why this matters

The failure DOS catches is the confident phantom: a write the agent authored and believes, contradicted by a witness the agent wrote zero bytes of. A booking "confirmed" that is not in the database. A fix: commit that only touched a README. A --allow-empty "implemented the cache" that changed nothing. Each one looks done. Each one is caught by reading the evidence, not the narration.

So: read the evidence. The tools are how.

Read the full file on GitHub · 57 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. 4d ago First seen · 57 lines · 606 tokens per session scan A 944f9f207014

Subscribe to this mod's changes

dos-kernel GEMINI.md is an instructions file published in the GitHub repository anthony-chaudhary/dos-kernel (19 stars, last pushed 10d ago), licensed MIT. It adds 606 tokens to every session, about $0.0030 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.