zeroshot: Instructions file for Codex

AGENTS.md

zeroshot AGENTS.md is an instructions file for Codex, OpenCode from the-open-engine/zeroshot. It costs 23,715 tokens per session, scanned A, original, MIT.

A set of repository instructions for Zeroshot, a tool that coordinates multiple software agents working on the same project.

In plain words
What is it for?
Use it to guide agents contributing to the Zeroshot codebase and to keep their work aligned with the repository’s development and release process.
Why use it?
It gives automated agents shared rules about permissions, branches, pull requests, validation, and project conventions.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md. Also seen: mentions CLAUDE.md; mentions Claude Code; runs codex exec.

This is the-open-engine/zeroshot's own configuration. It tells Codex and OpenCode how to work on zeroshot itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything zeroshot configures →

Needs its repository: it reads a path above its own folder, which exists only inside the repository. The line is cd ../...

About the project

Zeroshot is a software-change orchestration tool that separates implementation from verification: one coding agent makes a change and another independently checks whether it works. Developers use it to run coding tasks in isolated Git worktrees and continue the executor–verifier loop until the result is verified or a concrete failure is reported. The catalogue includes instructions, a hook, a skill, a setting, and a command for Zeroshot.

the-open-engine/zeroshot · 2,114 stars · on GitHub · zeroshot.sh

Reuse

Borrowing it

Nothing to install: this file belongs to the-open-engine/zeroshot. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/the-open-engine/zeroshot/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/the-open-engine/zeroshot

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 zeroshot AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/the-open-engine/zeroshot/agents-md/github.svg)](https://agentmods.dev/instructions/the-open-engine/zeroshot/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/the-open-engine/zeroshot/agents-md"><img src="https://agentmods.dev/badge/instructions/the-open-engine/zeroshot/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.

agentmods 80×15 button for zeroshot AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/the-open-engine/zeroshot/agents-md"><img src="https://agentmods.dev/badge/instructions/the-open-engine/zeroshot/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 23,715 This file is loaded in full into every session.
When invoked 23,715 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.23715 $0.23715
Opus 5.5 $0.09486 $0.09486
Sonnet 5.5 $0.04743 $0.04743
Haiku 4.5 $0.02371 $0.02371

Measured today against content hash 1d76d032eda7, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

zeroshot 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 · 1,337 lines

How it starts

The opening of the file, as written. The whole thing — 1,337 lines — stays where its author put it; the contents beside it link to each section on GitHub.

UPDATE THIS FILE when making architectural changes, adding patterns, or changing conventions.

Zeroshot v8

Operational guidance for automated agents working on this repository. Install the canonical command with npm i -g @the-open-engine-company/zeroshot or build zeroshot with Cargo.

Critical rules

  • Never run zeroshot run unless the user explicitly asks to start a run.
  • Never use git commands inside validator prompts; validators inspect files and observable outputs.
  • Agents are non-interactive. Make autonomous, scoped decisions rather than asking runtime questions.
  • Never edit CLAUDE.md unless the user explicitly requests it.
  • main is the only development and release trunk. Normal PRs target main.
  • Pull request titles are Conventional Commit headers because squash merge makes the title the released commit.
  • Pull request descriptions keep a nonempty ## Summary with the user-facing change. Release notes come from that summary in the immutable squash commit, not from mutable GitHub metadata.
  • Worker git operations are allowed only inside an isolated worktree/container or explicit PR/ship delivery flow.
  • Do not recreate the retired Node.js product, its commands, configuration, state, release workflow, package exports, compatibility aliases, migration, or dual publication identities.

Product and release identity

  • The Rust crate in zeroshot/ is the canonical product and owns the zeroshot CLI.
  • Node.js is used for repository tooling, the npm binary delivery package, and building the static profile UI. The UI is served by Rust and never needs a production Node server.
  • Canonical releases are explicit vX.Y.Z tags with major version 8 or newer.
  • The npm package is @the-open-engine-company/zeroshot.
  • That package owns the canonical zeroshot, zeroshot-memory-init, zeroshot-graph-design, and zeroshot-plan skills and installs managed copies for Codex, GitHub Copilot, Pi, and Claude Code at their user scopes. Do not fork skills by host. ZEROSHOT_SKIP_SKILLS=1 opts out of all user-scope skill handling in npm postinstall and zeroshot update; only the exact value 1 enables it. Executables still install/update, existing skills stay untouched, and updates skip skill downloads. Keep default managed-copy protections when the opt-out is unset.
  • The target image is ghcr.io/the-open-engine/zeroshot-target.
  • The Python distribution is the-open-engine-zeroshot; its import package remains zeroshot.
  • Python SDK tags are zeroshot-python-vZEROSHOT_SDK and package versions are ZEROSHOT.postSDK. Canonical releases publish SDK revision 1; later SDK revisions may release independently from an exact main commit descended from the canonical tag.
  • Canonical releases always publish revision 1 wheels to an immutable GitHub Release. publish_pypi defaults to true; operators may set it to false only when PyPI trusted publishing is known to be unavailable, then recover the same revision from the same source later.
  • Checked-in Cargo/npm versions are development placeholders. Tags, registry metadata, and GitHub Releases are authoritative. Never commit a staged release version to main.
  • zeroshot update resolves the newest canonical GitHub Release, verifies the declared host archive and canonical skill assets through that release's SHA256SUMS, smoke-checks the staged executable, and replaces the running executable in place without privilege escalation. It installs or refreshes only unchanged managed skill copies and preserves conflicts. Development-placeholder builds refuse self-update.
  • Release recovery may complete missing outputs only when existing immutable artifacts match the exact version and source commit.

Read the full file on GitHub · 1,337 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 · +64 lines · +1,200 tokens per session 1d76d032eda7
  2. yesterday Changed · +31 lines · +545 tokens per session 2777809e6e1a
  3. 3d ago Changed · +37 lines · +748 tokens per session f392ecd23b46
  4. 6d ago Changed c43b1a582b68
  5. 6d ago Changed · +97 lines · +1,717 tokens per session 953c8b1210d7
  6. 7d ago Changed · +32 lines · +591 tokens per session 5a6fcd8e6143
  7. 8d ago Changed · +17 lines · +327 tokens per session 2ad3894390ab
  8. 10d ago Changed · +4 lines · +77 tokens per session 106ee27b1a3c
  9. 12d ago Changed · +4 lines · +66 tokens per session b7f056853d0e
  10. 13d ago First seen · 1,051 lines · 18,444 tokens per session scan A 00d9c4bcd4d9

Subscribe to this mod's changes

zeroshot AGENTS.md is an instructions file published in the GitHub repository the-open-engine/zeroshot (2,114 stars, last pushed yesterday), licensed MIT. It adds 23,715 tokens to every session, about $0.0949 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 2026-08-30.

Related

zeroshot AGENTS.md compared with instructions from other repositories

zeroshot AGENTS.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
zeroshot AGENTS.mdthe-open-engine/zeroshotInstructions file2.1kyesterdayMITA23,715Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143ktodayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode