What is LLM-Wiki AGENTS.md?

Repository instructions for an Obsidian knowledge vault, where notes are linked and organized as a personal reference system.

Managing notes, using the vault’s search and writing commands, following Markdown conventions, and adding links and metadata consistently.

They explain which local rules and command instructions agents must follow when reading, finding, or writing vault content.

LLM-Wiki AGENTS.md — Instructions file by VectorPeak

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/vectorpeak/llm-wiki/agents-md

Clone the repo

git clone --depth 1 https://github.com/VectorPeak/LLM-Wiki

Made for: Codex, OpenCode.

Per session 3,809 This file is loaded in full into every session.

When invoked 3,809 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.

Model	Per session	Once invoked
Fable 5	$0.03809	$0.03809
Opus 5	$0.01904	$0.01904
Sonnet 5	$0.00762	$0.00762
Haiku 4.5	$0.00381	$0.00381

Measured 2d ago against content hash a656a775f8be, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

LLM-Wiki 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 2d 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 · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Obsidian Second Brain - Codex CLI Operating Manual

Local vault overrides: also follow _CLAUDE.md for folder conventions and Markdown authoring rules. In particular, ._trash&cache/ is reserved for temporary stitching/cache files and soft trash, and Mermaid diagrams in new Markdown should default to the forest theme with compact, whole-graph layouts.

This vault runs the obsidian-second-brain skill. The skill ships a set of commands: each one is a multi-step instruction file that you (the Codex agent) should follow when the user's request matches its trigger phrase.

How to operate

Read _CLAUDE.md in the vault root, if it exists, to learn the user's vault conventions (folder map, daily note format, naming).
When the user's request matches a trigger in the tables below, read the matching file under .codex/commands/<name>.md and follow its instructions step by step.
Treat the AI-first vault rule (.codex/references/ai-first-rules.md) as non-negotiable for every note you write: ## For future Claude preamble, rich frontmatter (type, date, tags, ai-first: true), [[wikilinks]] for every person/project/concept, recency markers per external claim, sources verbatim, confidence levels where applicable.
Do not invent commands. If no command matches, ask the user what they want or fall back to plain natural-language help.

Command routing tables (grouped by category)

Wiki layer - simplified knowledge operations

Command	What it does	Read this file
`/ingest`	LLM_wiki intake protocol - preserve raw sources, extract learning units, and update durable wiki knowledge	`02.wiki/projects/LLM_wiki Wiki Layer Protocol.md`, then `.codex/commands/ingest.md`
`/query`	LLM_wiki retrieval protocol - search, explain, synthesize, and answer from vault evidence	`02.wiki/projects/LLM_wiki Wiki Layer Protocol.md`, then `.codex/commands/query.md`
`/lint`	LLM_wiki maintenance protocol - audit links, duplicates, stale claims, raw backlog, DailyNotes drift, Mermaid defaults, and Mentor consistency	`02.wiki/projects/LLM_wiki Wiki Layer Protocol.md`, then `.codex/commands/lint.md`

Read the full file on GitHub · 180 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.

2d ago First seen · 180 lines · 3,809 tokens per session scan A a656a775f8be

Subscribe to this mod's changes

LLM-Wiki AGENTS.md is an instructions file published in the GitHub repository VectorPeak/LLM-Wiki (24 stars, last pushed 2mo ago), licensed MIT. It adds 3,809 tokens to every session, about $0.0190 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 2026-08-30.

LLM Memory and context agent ai-engineering deep-learning fine-tuning knowledge-base

LLM-Wiki AGENTS.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Obsidian Second Brain - Codex CLI Operating Manual

How to operate

Command routing tables (grouped by category)

Wiki layer - simplified knowledge operations

What this file has done since we first saw it

Other instructions, from other repositories

MindOS AGENTS.md

doco AGENTS.md

akb AGENTS.md

akb CLAUDE.md

ai-memex-cli AGENTS.md

ai-memex-cli CLAUDE.md