coreyhaines31/makerskills

makerskills – AI‑powered personal‑operator toolkit

What it is – A Claude Code (and other LLM‑agent) plugin that bundles 22 ready‑to‑run “skills”. Each skill is a small, documented workflow (e.g. decision‑making, domain‑name hunting, research, personal‑CFO modelling) that can be invoked from a chat interface with a slash command like /decide or /paste. The plugin ships the skill definitions as Markdown files (SKILL.md) and a thin runtime that parses the trigger, runs the appropriate LLM calls, and writes structured results to a user‑controlled config folder.

Who it’s for – Solo founders, indie makers, and small teams who already use Claude Code, Codex, Cursor, or similar AI‑agent platforms and want a library of repeatable, AI‑augmented workflows for everyday operator tasks.

Key capabilities

Area	Example skill	What it does
Decision & strategy	decide, unstuck, business‑brainstorm	Guides you through structured decision frameworks, surfaces hidden assumptions, and pressure‑tests ideas using the 37signals method and multi‑step research.
Knowledge management	second‑brain, company‑brain, radar	Captures, indexes, and queries personal or team markdown vaults; daily surveillance of feeds; integrates with tools like Fathom, Gong, or CRMs.
Content creation	read‑book, watch‑video, jab‑hook, slide‑deck	Extracts notes or transcripts, formats social‑media posts, builds slide decks from LLM‑generated outlines.
Operations & finance	pm, personal‑cfo, company‑cfo	Kanban/Eisenhower task boards, cash‑flow snapshots, scenario modelling, and automated data pulls via toolify.
Integration & automation	skillify, toolify, loopify	Wizards to create new skills, wire up APIs (e.g., Domainr, Namecheap), and schedule recurring agent loops.

How it works

1. Installation – Add the plugin via Claude Code’s marketplace or symlink the repo locally.
2. Configuration – Set MAKERSKILLS_CONFIG (and optional vault paths) in your shell; the plugin stores all generated artefacts under this directory.
3. Invocation – Type a slash command; the runtime reads the corresponding SKILL.md, runs the defined LLM prompts, may call external tools/APIs, and writes the output (often a markdown file with citations, dates, or structured JSON).
4. Composition – Skills can call each other (e.g., watch‑video → second‑brain → deep‑research), enabling complex pipelines without duplicating logic.

Installation snippet

# Claude Code marketplace
/plugin marketplace add coreyhaines31/makerskills
/plugin install makerskills@makerskills

# or local dev
git clone https://github.com/coreyhaines31/makerskills ~/code/makerskills
ln -s ~/code/makerskills ~/.claude/plugins/makerskills

Documentation style – Each skill lives in skills/<name>/SKILL.md with a YAML front‑matter block (name, description, metadata.version). The file doubles as user‑readable documentation and as the machine‑read trigger spec for Claude.

Ecosystem – The repo is open‑source (MIT) and references a sibling pack marketingskills for marketing‑focused workflows. Versioning is two‑tier: a global plugin tag and per‑skill semver, allowing independent iteration of individual workflows.

Why it matters – It turns a generic LLM agent into a personal operations hub, letting founders automate repetitive knowledge‑work (research, note‑taking, financial modelling) with minimal coding. The “documentation‑first, automation‑second” approach also makes the workflows transparent and easy to adapt.

---

All details are taken directly from the repository’s README; no external assumptions are added.