Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/supachai-j/open-knowledge-format-starternpx agentmods add skills/supachai-j/open-knowledge-format-starter/okfWrote 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/skills/supachai-j/open-knowledge-format-starter/okf)<a href="https://agentmods.dev/skills/supachai-j/open-knowledge-format-starter/okf"><img src="https://agentmods.dev/badge/skills/supachai-j/open-knowledge-format-starter/okf.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 5.1 | $0.00150 | $0.01272 |
| Opus 5 | $0.00075 | $0.00636 |
| Sonnet 5 | $0.00030 | $0.00254 |
| Haiku 4.5 | $0.00015 | $0.00127 |
Grade A, and why
okf 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 6d 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OKF Bundle Operator
This repo is an Open Knowledge Format (OKF) v0.1 bundle (Google Cloud, 2026-06-12): a directory of Markdown + YAML-frontmatter files that humans and agents co-maintain. You are the disciplined maintainer of this wiki, not a generic chatbot.
Read AGENTS.md at the repo root first — it is the authoritative schema. This skill is the
operating procedure layered on top of it.
Layout (3 layers)
raw/— immutable sources. Read only. Never edit.wiki/— the OKF bundle you own. Bundle root =wiki/, sowiki/tables/orders.md→ Concept IDtables/orders.tools/—concept-template.mdandokf-validate.py.- Reserved files:
wiki/index.md(catalog),wiki/log.md(append-only, newest-first).
Frontmatter rules (conformance)
- Required:
type(non-empty). Recommended:title,description,resource,tags,timestamp(ISO 8601 UTC). - Keep
typevalues within the controlled vocabulary inAGENTS.md. Inconsistent typing breaks aggregation. descriptionis what an agent reads to decide whether to load the file — make it a sharp one-liner.- Prefer structural Markdown (headings, atomic bullets, tables) over prose. Conventional body order:
# Overview→# Schema→# Common query patterns→# Joins→# Citations. - Link concepts with file-relative Markdown links (
../tables/customers.md). Never start a link with/— it breaks GitHub rendering (Google's reference agent forbids it). Links are untyped — relationship meaning lives in the prose. Only link to existing concepts; broken links are allowed (= unwritten knowledge). Don't over-link or link inside code/headings. - Curated/derived knowledge (joins, metric defs, glossary) goes under
references/withtype: Reference; tangible assets undertables//datasets/.
Operation: INGEST (supervised — this is a quality gate, never automatic)
- Read the source in
raw/. - Read
wiki/index.mdto see what concepts already exist. - Extract 5–15 key claims/decisions worth keeping.
- Show the user the extracted claims + proposed concept mappings, and WAIT for approval before writing.
- Map claims to existing concepts; create a NEW concept only if 5+ claims share an uncovered theme.
- Reconcile contradictions: if new data conflicts with an existing concept, append to the OLD file
> **CONTRADICTION FLAG**: <explanation; see <new concept ID>>and add context to the new one. - Write/update concept files (start from
tools/concept-template.md); updatetags+timestamp. - Update the affected
index.mdfiles. - Append one entry to
wiki/log.mdunder today's ISO date heading:- ingest | <source title> | touched: <concept ids>. - Run VALIDATE (below) before declaring done.
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.
- 6d ago First seen · 79 lines · 150 tokens per session scan A 44b84b6c0ce2
okf is a skill published in the GitHub repository supachai-j/open-knowledge-format-starter (5 stars, last pushed 2mo ago), licensed MIT. It adds 150 tokens to every session and 1,272 once invoked, about $0.0007 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 skills, from other repositories
llm-wiki
Build and maintain a persistent, interlinked Obsidian-compatible markdown wiki using Karpathy's LLM Wiki pattern. Extension-backed with auto-generated metadata, guardrails, and 14 custom tools (+3 opt-in agent-trajectory tools).
link-memory
Use after important user-approved decisions, when durable context should be proposed or reviewed, and for explicit Link memory lifecycle work: remember, recall, review, update, archive, restore, forget, or explain local memories through the CLI without requiring MCP.
link-retrieve
Use before answering work that may depend on user memory, project history, source-backed notes, or prior decisions; retrieve compact Link context through the CLI without loading the whole wiki or requiring MCP.
link-health
Use at the start of Link work when readiness is unclear, after installs or upgrades, and before repairs; verify health, inspect interrupted writes, back up, and repair generated indexes without MCP.
link-ingest
Use when raw files are present, source pages look stale, or a user asks to ingest notes into Link; refresh source-backed wiki pages, propose memories, and validate updates through the CLI without MCP.
mindbase
Maintain a persistent AI-tended research wiki. Activates when the user wants to ingest a source (PDF, URL, article, paste), capture a thought, build the wiki context, query what's already known, lint for orphans/contradictions, or maintain page conventions. Phrases include "add this to my wiki", "today I decided…"…