Free playbooks in your inbox
Home / Series / AI Memory Playbooks
Collection

AI Memory Playbooks.

A file layer your agent actually reads (Obsidian vaults, CLAUDE.md schemas, HTML artifacts, LLM wikis) and a council of advisors that argues before it agrees. Stop re-explaining your project every session, and stop trusting a single yes-man model.

4 titles · 20 free guides

Free guides
tutorial 9 min read

Five Claude Code Lifecycle Hooks for Your Obsidian Vault

Configure five Claude Code lifecycle hooks (SessionStart, UserPromptSubmit, PostToolUse, PreCompact, Stop) that load and re-inject vault context every prompt.

from: Claude Code + Obsidian: The AI Second Brain Playbook

tutorial 8 min read

How to Build a 5-Folder Obsidian Vault for Claude Code

Ship the Obsidian vault structure Claude Code uses as primary context: five folders, install commands, and the smoke test that proves the loop in 60 minutes.

from: Claude Code + Obsidian: The AI Second Brain Playbook

tutorial 8 min read

Three Claude Code Subagents for an Obsidian Vault

Three Claude Code subagents (brag-spotter, slack-archaeologist, cross-linker) for an Obsidian vault: isolated contexts, three Friday runs, fifteen minutes.

from: Claude Code + Obsidian: The AI Second Brain Playbook

how-to 8 min read

The 7-Section CLAUDE.md Template for an Obsidian Vault

A CLAUDE.md Obsidian vault template that survives long sessions: identity, folder schema, active projects, decision log, anti-patterns, sized 80-150 lines.

from: Claude Code + Obsidian: The AI Second Brain Playbook

tutorial 8 min read

Local Semantic Search for an Obsidian Vault with Ollama

Set up obsidian semantic search with Ollama and obsidian-copilot in 20 minutes. Local embeddings, no data leaves your machine, three-tier retrieval ladder.

from: Claude Code + Obsidian: The AI Second Brain Playbook

tutorial 9 min read

Build a Throwaway Claude Prompt Tuner in One Prompt

Build a claude prompt tuner as one HTML file: template with slot highlighting, three live-rendering examples, token counter. Thariq's verbatim prompt + fix.

from: Files Over Prompts

how-to 9 min read

Ship HTML Artifacts on Every PR (Claude Code Pattern)

Html artifacts claude engineers attach to every PR. The verbatim Thariq prompt, three first-run failure modes, and the 20% surfaced first that beats markdown.

from: Files Over Prompts

tutorial 9 min read

How to Build Karpathy's LLM Wiki in 90 Minutes

The karpathy llm wiki is two folders and a schema file, not three folders. Build the real structure on disk in 90 minutes with the mkdir commands and 5 entries.

from: Files Over Prompts

analysis 9 min read

LLM Wiki Maintenance: One Wrong Summary, 16 Errors

Llm wiki maintenance fails when one bad summary cascades. The 1 to 4 to 16 error pattern, the lint pass that cuts the tree, and the cost math at every scale.

from: Files Over Prompts

how-to

How to Charge Monthly for an Obsidian Vault You Set Up Once

The install is two hours and a free video gives it away, so the install is not the business. Here is the one-page scope that makes a monthly fee renew, and the audit you sell first so you can quote the engagement instead of guessing it.

from: Obsidian + Claude Code Build the Dreaming Loop

how-to

How to Stop Claude Code From Editing Your Important Files

Point an agent at a directory you cannot replace and know, because you watched a tool call get blocked, that it cannot touch the part that matters. A path rule written against Write is accepted at startup and ignored forever, and the fix is one word plus a hook that exits 2.

from: Obsidian + Claude Code Build the Dreaming Loop

how-to

Did Your Claude Code Agent Actually Do Anything Last Night?

Put real work on a schedule and wake up genuinely knowing what happened, because the job greps its own exit signal and tells you. Eleven consecutive green runs, and every one was four seconds of a job that could not find its input file.

from: Obsidian + Claude Code Build the Dreaming Loop

tutorial

Which of Your Obsidian Notes Is Claude Getting Wrong?

Run one read-only pass and you can hand somebody a dated document naming the pages your vault is already wrong about. Seventeen findings across two hundred and fourteen pages, each one citing the two files that disagree and the typed field they disagree on.

from: Obsidian + Claude Code Build the Dreaming Loop

analysis

Why Claude Still Answers From the Obsidian Note You Replaced

Twenty questions from your own history end an argument four credible people could not, because it is the only evidence in the room about your data. The arm that wins on raw recall is also the one that hands back the fact you retired.

from: Obsidian + Claude Code Build the Dreaming Loop

how-to 7 min read

Build an AI Board of Advisors That Argues Back

Stop making pricing and hiring calls with a yes-man AI. Build an AI board of advisors: four roles that pressure-test the decision and return one honest verdict.

from: The LLM Council

tutorial 8 min read

AI Code Review with a Multi-Model Council

Make Codex, Claude, and Gemini argue over your diff and block on any single real bug. Build a multi-model AI code review council that catches what one misses.

from: The LLM Council

tutorial 8 min read

How to Build an LLM Council in Claude Code

Build a five-advisor LLM council as a Claude Code skill that runs five independent AI advisors on one decision, in one command, with no two-terminal setup.

from: The LLM Council

tutorial 8 min read

Build a Multi-Model AI Council with OpenRouter

Route one question to GPT, Claude, Gemini, and Grok through OpenRouter so no single vendor's blind spot decides your answer. The multi-model council build.

from: The LLM Council

how-to 7 min read

How to Verify AI Claims with a Council

Fact-gate a claim with a four-check AI council that returns SUPPORTED, CONTRADICTED, or an honest UNVERIFIABLE. How to verify AI claims instead of trusting one.

from: The LLM Council

The series, in order

Read them in this order. Each one assumes the one before it.

01 Book
Claude Code + Obsidian: The AI Second Brain Playbook

Claude Code + Obsidian: The AI Second Brain Playbook

A vault Claude actually reads: 5 folders, 5 hooks, 3 subagents

02 Book
Files Over Prompts

Files Over Prompts

A 5-entry LLM wiki, persistent memory, and three throwaway HTML editors in one weekend

03 Book
The LLM Council

The LLM Council

Five advisors that answer independently, peer-review anonymously, and return one verdict with the dissent named

04 Book
Obsidian + Claude Code Build the Dreaming Loop

Obsidian + Claude Code Build the Dreaming Loop

A scheduled loop that finds the rot, stages the fix for your approval, and proves recall with a golden set ninety days later

Want the buying order, not the shelf?

These titles sit on a path that tells you which three to buy first.