Skip to content

Skillselionwith God's help

By phase

IdeaResearch & spot demand
ValidateScope & prototype
BuildFrontend, backend, MCP
ShipTest, review, deploy
LaunchDistribution, SEO, ASO
GrowAnalytics & retention
OperateMonitoring & uptime

By type

Skills
Marketplaces
MCP servers
Workflows

Skills Marketplaces MCP Learn Brief

Browse all tools →By phase

Idea Validate Build Ship Launch Grow Operate

Explore

Skills
Marketplaces
MCP
Learn
Brief

This week for builders

Five minutes, every Monday — the tools, releases and tactics for shipping solo.

Email address

unsubscribe anytime.

Skillselion

An independent, curated directory of AI-coding tools for solo builders - skills, MCP servers, marketplaces and workflows for Claude Code, Codex and Cursor, organized by the indie build journey from idea to operate.

ResourcesBrowse Skills Browse MCP Servers Browse Marketplaces Browse Workflows Best-of shortlists Glossary State of AI Skills 2026 Weekly Brief

CommunityAbout Learn Feedback Privacy Policy Terms Cookie Policy Accessibility Cookie settings

Last updated June 10, 2026 · © 2026 Skillselion. Skillselion is an independent project and is not affiliated with Anthropic, OpenAI, Cursor, Claude, Claude Code, or Codex.

Skills/NASHSU/LLM_WIKI_SKILL

nashsu/llm_wiki_skill

1 skill · 3 installs · 59 stars · GitHub

Install

npx skills add https://github.com/nashsu/llm_wiki_skill

Skills in this repo

1Llm WikiLLM Wiki is an agent skill that talks to the user’s locally running LLM Wiki desktop application through its built-in HTTP API on localhost. Solo builders who maintain a curated wiki—markdown pages under wiki/, raw documents, and wikilinks forming a graph—install it so Claude Code, Cursor, or similar agents can answer questions grounded in that knowledge base instead of hallucinating context. Invoke it when the user clearly means LLM Wiki: named projects, “what does my wiki say about X”, graph views, or rescan/reindex requests. It deliberately does not replace other PKM tools; generic “search my notes” should route elsewhere. The workflow is direct REST-style calls with whatever HTTP client the environment already has, keeping permissions narrow except when the user asks to rescan sources. That makes it valuable during implementation when specs live in the wiki, during validation when prior research is stored there, and during operations when runbooks and architecture notes are in the same graph.3installs