skills.rest
daymade
daymadeCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

llm-wiki-setup

Scaffolds a markdown investment-research wiki vault and interviews users to author their own CLAUDE.md.

1.4k|216|Updated Oct 22, 2025
Search Tags:#investment research#wikilinks#markdown vault#earnings review#Knowledge Base#LLM Wiki#clauDE.md
One-click install
npx skills add https://github.com/daymade/claude-code-skills --skill llm-wiki-setup
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: llm-wiki-setup
Source: https://github.com/daymade/claude-code-skills/tree/main/llm-wiki-setup
Command: npx skills add https://github.com/daymade/claude-code-skills --skill llm-wiki-setup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Building a personal investment-research knowledge base usually means adopting a generic template that erases the investor's own judgment framework. This Skill co-creates a Karpathy-style LLM Wiki (pure markdown plus wikilinks, no RAG or vector database) where the user's own analysis preferences become a living CLAUDE.md, extracted through a structured interview rather than copied from a template.

Core Features & Use Cases

  • Vault Scaffolding: Runs init_vault.py to create the three-layer directory structure (raw/, wiki/, index, log), a lint script, and a git pre-commit hook that blocks broken wikilinks and invalid YAML.
  • Interview-Driven CLAUDE.md: Guides an 8-dimension interview (markets covered, decision signals, layering, analyst attribution, time views, output format, review cadence, source types) and writes answers in the user's own words.
  • Ingest and Review SOPs: Provides human-in-the-loop checkpoints for ingesting research reports and earnings calls, plus post-earnings prediction-versus-outcome fulfillment reviews.
  • Use Case: An investor tracking AI supply-chain stocks wants a compounding research vault. The Skill scaffolds the vault, interviews them to capture that they watch order momentum and management tone, then ingests their first sell-side report with confirmation checkpoints.

Quick Start

Ask the assistant to set up a personal investment-research LLM Wiki in a new folder and interview you to write your own CLAUDE.md.

Frequently Asked Questions about llm-wiki-setup

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a personal investment research knowledge base with markdown?▼

Run the init_vault.py script to scaffold a vault with raw/ and wiki/ directories, an index, and a log. Then complete the guided interview so your own decision criteria are written into CLAUDE.md, and ingest sources through the documented SOP with confirmation checkpoints.

What is Karpathy's LLM Wiki pattern for knowledge management?▼

It is a pattern of plain markdown files connected by wikilinks and maintained by an LLM, with knowledge pre-compiled into structured pages instead of retrieved per query. This Skill instantiates that pattern for finance and investing without adding RAG, vector databases, or embeddings.

Does this wiki use RAG or a vector database for retrieval?▼

No. The Skill deliberately avoids RAG, vector stores, and embeddings, treating pure markdown plus wikilinks and grep as the core of the pattern. Knowledge compounds through structured pages rather than re-retrieving raw documents at query time.

How does the vault lint script catch broken wikilinks?▼

The lint-vault.py script runs as a git pre-commit hook and hard-fails commits containing wikilinks to nonexistent pages, invalid YAML frontmatter, or company pages lacking industry or macro links. It also warns about orphaned raw files and oversized pages without blocking.

When should I not use a template for my investment wiki?▼

You should never copy a template's rule layer, because the wiki's value comes from encoding your own judgment criteria in your own words. The provided example CLAUDE.md is illustrative only; the interview process exists to extract your personal dimensions instead.

Related Skills

Explore More Finance & Accounting Skills→
byteseekbyteseek
252

Research Report Interpretation Skill

Parse investment research reports into verifiable claims and assumptions.

Official
Advanced
loanntcloanntc

earnings-forecast

Generate trading signals by forecasting EPS and computing SUE with PEAD logic.

Community
Advanced
werfwerf
4.7k

inherited-claims

Re-verify conclusions inherited from prior sessions against primary sources before publishing.

Official
Basic
jhl-labsjhl-labs
50

RAG & Vector Search

Retrieve relevant documents and augment LLM prompts with vector search.

Official
Advanced