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.
npx agentmods add instructions/supachai-j/open-knowledge-format-starter/agents-mdgit clone --depth 1 https://github.com/supachai-j/open-knowledge-format-starterWrote 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.
[](https://agentmods.dev/instructions/supachai-j/open-knowledge-format-starter/agents-md)<a href="https://agentmods.dev/instructions/supachai-j/open-knowledge-format-starter/agents-md"><img src="https://agentmods.dev/badge/instructions/supachai-j/open-knowledge-format-starter/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01843 | $0.01843 |
| Opus 5 | $0.00922 | $0.00922 |
| Sonnet 5 | $0.00369 | $0.00369 |
| Haiku 4.5 | $0.00184 | $0.00184 |
Grade A, and why
open-knowledge-format-starter 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.
How it starts
The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — OKF Bundle Schema & Agent Operating Rules
This is the schema file for this repository. It is the first file an agent reads. It tells the agent how the knowledge base is structured and how to behave when ingesting sources, answering questions, or maintaining the wiki. (Equivalent file names for other agents:
CLAUDE.md,GEMINI.md.)
This repository follows the Open Knowledge Format (OKF) v0.1 (Google Cloud,
2026-06-12) layered on the "LLM-wiki" pattern. The OKF bundle lives in wiki/.
1. Directory layout (3-layer architecture)
/ (repo root)
├── AGENTS.md ← this schema (agent behavior + conventions)
├── raw/ ← LAYER 1: immutable source materials. Agent READS, never edits.
├── wiki/ ← LAYER 2: the OKF bundle. Agent OWNS and maintains this.
│ ├── index.md ← reserved: progressive-disclosure catalog
│ ├── log.md ← reserved: append-only chronological audit trail
│ ├── tables/
│ ├── metrics/
│ └── playbooks/
└── tools/ ← scripts + the concept template
raw/is the source of truth. Drop PDFs, notes, exports here. Read-only for the agent.wiki/is the OKF bundle. Its root iswiki/, so a file atwiki/tables/orders.mdhas Concept IDtables/orders(path within the bundle, minus.md).tools/holds helpers:concept-template.md,okf-validate.py(conformance checker), andokf-viz.py(generateswiki/viz.html, a self-contained interactive graph viewer — a port of the visualizer from Google's reference implementation).
2. Concept files & frontmatter schema
Every non-reserved .md file in wiki/ is a Concept — one unit of knowledge.
Required (OKF v0.1 conformance):
type— short string for the kind of concept. Use the controlled vocabulary below.
Recommended (priority order):
title— human-readable name.description— one-line relevance summary. This is what an agent reads to decide whether to load the file — write it well.resource— URI of the underlying real-world asset (omit for self-describing curated concepts).tags— YAML array, e.g.[sales, revenue].timestamp— ISO 8601 UTC of the last meaningful change.
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.
- 4d ago First seen · 145 lines · 1,843 tokens per session scan A 5b8793619a75
open-knowledge-format-starter AGENTS.md is an instructions file published in the GitHub repository supachai-j/open-knowledge-format-starter (5 stars, last pushed 2mo ago), licensed MIT. It adds 1,843 tokens to every session, about $0.0092 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.
Other instructions, from other repositories
robust-llm-wiki CLAUDE.md
Instructions for KevinYoung-Kw/robust-llm-wiki, covering claude.md, the original karpathy baseline, made explicit and repository-specific policy.
agent-wiki CLAUDE.md
Instructions for xccElephant/agent-wiki, covering llm wiki agent guide, directory roles, core rules, claim types and ingest workflow.
llm-wiki-agent AGENTS.md
AGENTS.md instructions for SamurAIGPT/llm-wiki-agent, covering llm wiki agent — schema & workflow instructions, how to use, directory layout, page format and ingest workflow.
llm-wiki-agent GEMINI.md
Gemini CLI instructions for SamurAIGPT/llm-wiki-agent, covering llm wiki agent — schema & workflow instructions, how to use, directory layout, page format and ingest workflow.
llm-wiki-compiler AGENTS.md
AGENTS.md instructions for atomicstrata/llm-wiki-compiler, covering llmwiki, development guidelines, code style & standards, clean code rules and comments and documentation.
pi-llm-wiki AGENTS.md
AGENTS.md instructions for zosmaai/pi-llm-wiki, covering agents.md, project, tech stack, file layout and conventions.