token-station AGENTS.md

token-station AGENTS.md is an instructions file for Codex, OpenCode from ballast-ai/token-station. It costs 3,062 tokens per session, scanned A, original, Apache-2.0.

A project instruction file for ballast-ai/token-station that sets rules for writing, committing, and documenting code.

In plain words
What is it for?
Keeping technical content in English, writing the design document first, and committing authorized changes directly to the main branch.
Why use it?
It gives coding agents consistent project rules, so they do not guess about language, branch use, commit style, or design notes.

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/ballast-ai/token-station/agents-md
Clone the repo
git clone --depth 1 https://github.com/ballast-ai/token-station

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 token-station AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/ballast-ai/token-station/agents-md.svg)](https://agentmods.dev/instructions/ballast-ai/token-station/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/ballast-ai/token-station/agents-md"><img src="https://agentmods.dev/badge/instructions/ballast-ai/token-station/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,062 This file is loaded in full into every session.
When invoked 3,062 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 $0.03062 $0.03062
Opus 5 $0.01531 $0.01531
Sonnet 5 $0.00612 $0.00612
Haiku 4.5 $0.00306 $0.00306

Measured 4d ago against content hash d939021874b3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

token-station 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 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.

AGENTS.md · 288 lines

How it starts

The opening of the file, as written. The whole thing — 288 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Token Station Engineering Rules

Use English by default

Use English for new and modified technical content. This rule applies to code comments, documentation, design records, test names, commit messages, pull requests, release notes, and contributor instructions.

Use ASCII English words for technical-document filenames. Locale suffixes such as zh-CN are allowed, but localized words are not.

Keep commands, paths, protocol fields, API names, and code identifiers unchanged. Keep required localized user-interface text in its target language. You can quote an error or log entry in its original language.

If a document must contain English and Simplified Chinese, put the complete English text first. Put the complete Simplified Chinese text after it.

Write commit messages in English. Use a short imperative subject. Describe one logical change in each commit.

Commit directly to main

Work on the main branch. Commit authorized changes directly to main.

Do not create feature, topic, release, or migration branches unless the user explicitly requests one. Do not create a pull request unless the user explicitly requests one.

After you complete a feature or task, commit all authorized changes locally. Use an English commit message.

Before each commit, verify that the current branch is main. Stage only the files authorized for that commit. Keep unrelated working-tree changes unstaged.

Use Simplified Technical English

Use strict Simplified Technical English for procedures, runbooks, safety warnings, and error messages.

  • Put one instruction in each sentence.
  • Put the condition before the action.
  • Use the imperative form for instructions.
  • Keep an instruction at 20 words or fewer when practical.
  • Keep a descriptive sentence at 25 words or fewer when practical.
  • Use active voice when the actor is known.
  • Do not use contractions or semicolons.
  • Use one stable name for one thing.
  • Prefer short and common words.

Use STE-flavored English for READMEs, design documents, pull requests, release notes, and general technical explanations. Keep the text direct and natural. Remove filler, marketing claims, and unnecessary abstract terms.

Read the full file on GitHub · 288 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 · 288 lines · 3,062 tokens per session scan A d939021874b3

Subscribe to this mod's changes

token-station AGENTS.md is an instructions file published in the GitHub repository ballast-ai/token-station (7 stars, last pushed 7d ago), licensed Apache-2.0. It adds 3,062 tokens to every session, about $0.0153 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 2026-08-31.