ardur-agent AGENTS.md

ardur-agent AGENTS.md is an instructions file for Codex, OpenCode from ArdurAI/ardur-agent. It costs 1,210 tokens per session, scanned A, original, Apache-2.0.

A session guide for Ardur Agent, including a required read-only startup check and rules for finding and claiming work.

In plain words
What is it for?
Use it before planning, editing, testing, or handing off work in the ardur-agent project.
Why use it?
It prevents agents from working on the wrong task or interfering with parallel sessions by defining which project records and files are authoritative.

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/ardurai/ardur-agent/agents-md
Clone the repo
git clone --depth 1 https://github.com/ArdurAI/ardur-agent

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 ardur-agent AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/ardurai/ardur-agent/agents-md.svg)](https://agentmods.dev/instructions/ardurai/ardur-agent/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/ardurai/ardur-agent/agents-md"><img src="https://agentmods.dev/badge/instructions/ardurai/ardur-agent/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,210 This file is loaded in full into every session.
When invoked 1,210 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 $0.01210 $0.01210
Opus 5 $0.00605 $0.00605
Sonnet 5 $0.00242 $0.00242
Haiku 4.5 $0.00121 $0.00121

Measured 3d ago against content hash ee1cddc86322, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ardur-agent 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 3d 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 · 143 lines

How it starts

The opening of the file, as written. The whole thing — 143 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Ardur Agent Session Instructions

These instructions apply to every LLM, coding agent, and human-assisted agent session that works in this repository.

Mandatory Bootstrap

Before planning, editing, testing, or creating a handoff, run:

python3 scripts/agent_bootstrap.py

Use --json only when a tool needs machine-readable output:

python3 scripts/agent_bootstrap.py --json

The bootstrap is read-only. It reports current Linear status, project progress, Git/worktree state, provider and memory posture, unowned parallel:ready candidates, the Plan Corpus verification backlog, and the required non- interference rules for parallel sessions.

If the generic Linear connector shows a workspace or team other than ardur-agent / ARD, do not claim that work. The bootstrap uses the Keychain-backed helper at architect/tools/linear_graphql.py and must be the source for ARD work discovery.

Source Hierarchy

  1. Linear ARD is the internal source of truth for work scope, active state, priority, acceptance criteria, progress, blockers, and verification evidence.
  2. EXTENDED drive is the primary file/evidence source for plans, journals, audits, local branches, and implementation evidence.
  3. GitHub ArdurAI/ardur-agent is the verified public code surface for PRs, CI, releases, and public collaboration.
  4. Notion is a knowledge projection, useful for durable explanation and navigation, but it is not the authority for whether work is complete.

Session Ownership Rule

Before code edits, a session must claim one unowned parallel:ready Linear issue. After claiming it, update Linear to parallel:owned or In Progress and record:

  • one Linear issue,
  • one branch or isolated worktree,
  • one EXTENDED-drive session journal,
  • one explicit file/module scope.

Never claim issues from a non-ARD workspace while working in this repository. Wrong-workspace Linear access is a blocker, not a reason to switch projects.

Preferred branch format:

Read the full file on GitHub · 143 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. 3d ago First seen · 143 lines · 1,210 tokens per session scan A ee1cddc86322

Subscribe to this mod's changes

ardur-agent AGENTS.md is an instructions file published in the GitHub repository ArdurAI/ardur-agent (2 stars, last pushed 7d ago), licensed Apache-2.0. It adds 1,210 tokens to every session, about $0.0060 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.