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/green-dalii/obsidian-llm-wiki/agents-mdgit clone --depth 1 https://github.com/green-dalii/obsidian-llm-wikiWhat 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 | $0.07700 | $0.07700 |
| Opus 5 | $0.03850 | $0.03850 |
| Sonnet 5 | $0.01540 | $0.01540 |
| Haiku 4.5 | $0.00770 | $0.00770 |
Grade A, and why
obsidian-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 yesterday.
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 — 393 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LLM Wiki Plugin Project Development Standards
Last Updated: 2026-08-27
Current state pointer: see ROADMAP.md for active development phase + CHANGELOG.md for shipped history. This file (AGENTS.md) carries process standards only — and is the single canonical source. The historical mirror at CLAUDE.md is now a pointer stub to this file; all new content goes here. Latest shipped: v1.27.0 MINOR (2026-08-27, 3677 tests / 260 files).
🛡️ Six-Gate Quality Closure
| Gate | Constraint | How |
|---|---|---|
| 1. Code correct | pnpm lint 0/0 + npx tsc --noEmit 0/0 + pnpm build clean + pnpm test all pass + pnpm css-lint 0 |
Five-Gate script (build BEFORE test — see §"Gate 1: Five-Gate automated") |
| 2. No side effects | Call-site audit + data flow + state mutation + error propagation | Structured review |
| 3. No breaking changes | API/Schema/File format/Default behavior/Command IDs/Obsidian API | Breaking-change matrix |
| 4. No performance regression | CPU/memory/IO/network/token — 5-dim written assessment | simplify + code-review + Gate 4 table |
| 5. Docs complete | 10 READMEs + ROADMAP + AGENTS + CHANGELOG + memory | pre-release-gate |
| 6. Release clean | Supersets 1-5 + TOC + i18n + Release Notes + Contributors + git hygiene | pre-release-gate |
Gate 1: Five-Gate automated
pnpm lint && npx tsc --noEmit && pnpm build && pnpm test && pnpm css-lint
All five must pass. ESLint checks style, TypeScript checks types, css-lint checks Obsidian review compliance — three complementary checks, single tool passing is insufficient. No @ts-ignore / eslint-disable to silence failures.
Order is non-negotiable: pnpm build MUST run before pnpm test. The test suite contains build-artifact verifications (e.g. src/__tests__/llm-sdk/openai-codex-loopback-flow.test.ts:39 reads main.js to assert the esbuild bundle shape), so a test-first run fails with ENOENT on a fresh clone. Local Gate 1 typically has main.js on disk from a prior pnpm build:dev, which is why this ordering bug was missed before PR #487's first CI run on 2026-08-18.
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.
- yesterday First seen · 393 lines · 7,700 tokens per session scan A 75374c1cc5f5
obsidian-llm-wiki AGENTS.md is an instructions file published in the GitHub repository green-dalii/obsidian-llm-wiki (515 stars, last pushed 2d ago), licensed Apache-2.0. It adds 7,700 tokens to every session, about $0.0385 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.
Other instructions, from other repositories
pi-llm-wiki AGENTS.md
Instructions for zosmaai/pi-llm-wiki, covering agents.md, project, tech stack, file layout and conventions.
llm-brain CLAUDE.md
Instructions for kimsanguine/llm-brain, covering llm-brain — claude code 운영 지침, 가드레일 (절대 위반 금지), 프라이버시 경계 (agent memory os), 명령어 and ingest.
zettelkasten-second-memory AGENTS.md
Instructions for cx2002302-lang/zettelkasten-second-memory, covering zettelkasten project guidelines, 系统版本信息（每次思考必须考虑）, 兼容性保障索引, openclaw 2026.6.x 兼容性变更 and openclaw 2026.4.24 配置字段备忘.
obsidian-mcp-router AGENTS.md
AGENTS.md instructions for tboome33/obsidian-mcp-router, covering agents.md — operating contract for obsidian-mcp-router, what this repository is, the three gates, contracts — the parts that fail loudly on drift and using the skills (the bridge rule).
commonplace CLAUDE.md
Instructions for briannaworkman/commonplace: This repository holds two vaults. Nothing should be ingested or edited here at the root.
cmds-llm-wiki AGENTS.md
Schema and harness document for the CMDS LLM Wiki vault. Defines the 3-layer architecture (Raw Sources / Wiki / Schema), ingest-query-lint operations, file conventions, and frontmatter standards. This is the single source of truth for LLM behavior in this vault.