Makerskills - Setup Guide¶
Source: coreyhaines31/makerskills GitHub: coreyhaines31/makerskills (681⭐, created June 3, 2026, actively maintained) Catalog size: 23 skills, 4.7K total installs (publisher page, authoritative count) Category: Business Operations & Knowledge Management First Seen: July 2, 2026 on skills.sh Quality Tier: 🟡 Trusted (Gen Agent Trust Hub: Pass, Socket: Pass, Snyk: Warn)
Makerskills is a cohesive suite of agent skills for the personal operator's craft: decisions, research, knowledge vaults, content rotation, scenario modeling, and finance. Each skill is a standalone SKILL.md playbook with standard frontmatter (name, description, metadata) and no platform binding, so they load natively in Hermes Agent like any other skill.
Compatibility note: the repo README targets Claude Code, Codex, and Cursor, and the playbook prose is Claude-flavored ("so Claude can answer questions"). The mechanics are agent-agnostic procedural workflows (vaults, modes, triggers) with no Claude-specific tool requirements; Hermes users install through the skills.sh CLI or by copying individual SKILL.md files.
Overview¶
| Skill | Installs | Domain |
|---|---|---|
watch-video |
339 | Content intake |
decide |
303 | Decision workflow |
social-fetch |
268 | Social content intake |
read-book |
265 | Reading workflow |
business-brainstorm |
242 | Strategy |
second-brain |
241 | Personal knowledge vault |
jab-hook |
239 | Content rotation |
company-brain |
238 | Team knowledge vault |
slide-deck |
238 | Presentation generation |
loopify |
231 | Workflow automation |
toolify |
231 | Tool authoring |
pm |
231 | Product management |
paste |
228 | Clipboard sharing |
skillify |
227 | Skill authoring |
deep-research |
223 | Multi-source research |
company-cfo |
221 | Business finance |
domain |
216 | Domain research |
personal-cfo |
216 | Personal finance |
unstuck |
167 | Creative unblocking |
maker-council |
96 | Scenario modeling |
adapt-skill |
1 | Meta: adapt skills |
update-skill |
1 | Meta: update skills |
create-skill |
1 | Meta: author skills |
Not included: none excluded; the three 1-install meta-skills (adapt-skill, update-skill, create-skill) are authoring utilities documented in the roster above rather than given separate guides.
Installation¶
Install the full suite via the skills.sh CLI:
npx skills add coreyhaines31/makerskills
Install individual skills by name:
npx skills add https://github.com/coreyhaines31/makerskills --skill company-brain
Or copy a playbook directly into your Hermes skills directory (the repo stores 20 SKILL.md files under skills/):
mkdir -p ~/.hermes/skills/makerskills
curl -sL https://raw.githubusercontent.com/coreyhaines31/makerskills/main/skills/company-brain/SKILL.md \
-o ~/.hermes/skills/makerskills/company-brain/SKILL.md
Prerequisites¶
| Requirement | Details |
|---|---|
| Hermes Agent | Any version with skills directory support (~/.hermes/skills/) |
| Node.js 18+ | Only for the npx skills CLI path |
| No API keys | The playbooks are knowledge and decision guides; they use tools you already have configured |
| Network (optional) | Only needed for npx skills fetch or raw.githubusercontent.com downloads |
What It Provides¶
Knowledge vaults¶
second-brain wraps the Karpathy LLM Wiki schema over any Obsidian or markdown vault with a raw → wiki → outputs pipeline (capture, compile, query, lint, connect, search modes; defaults to ${SECOND_BRAIN_VAULT:-$HOME/Documents/SecondBrain}). company-brain is the team-scope sibling with structured category dirs (people, companies, meetings, SOPs, decisions, sales-objections) plus capture, compile, query, review, lint, connect, and search modes, with author and timestamp stamping on every capture.
Research and decisions¶
deep-research runs multi-pass research with source planning, verification, citations, contradictions, and gap analysis, archiving every run to ~/.config/makerskills/deep-research/archive/ so the corpus compounds. decide walks the 37signals Guide to Making Decisions question set (plus an opportunity-cost question), triages to the 6-8 load-bearing questions, and archives every call with a revisit date.
Business operations¶
company-cfo and personal-cfo bring finance discipline to business and personal money questions. business-brainstorm and maker-council cover strategy generation and scenario modeling. pm applies product-management discipline to spec work. domain researches a business domain before diving in.
Content and workflow¶
watch-video, read-book, and social-fetch turn content into vault captures. jab-hook manages content rotation, slide-deck generates presentations from vault material, paste shares clipboard state, unstuck breaks creative blocks, and loopify chains skills into repeatable workflows. skillify, toolify, and the three meta-skills close the loop by authoring new skills from your own patterns.
Quick Start¶
# Install the full suite
npx skills add coreyhaines31/makerskills
# Start a personal knowledge vault
# Trigger: "/second-brain" or "capture this"
# Start a team knowledge vault
# Trigger: "/company-brain" or "log this meeting"
# Run a research pass
# Trigger: "/deep-research" or "do a deep dive on X"
CorpusIQ Use Cases¶
| Use Case | How |
|---|---|
| Memory architecture | Use second-brain and company-brain patterns as reference designs alongside GBrain and Honcho when discussing agent memory layers |
| Lead research | Use deep-research for pre-call prospect research and due diligence |
| Deal decisions | Use decide to log go/no-go calls with revisit dates on partnerships and placements |
| Finance checks | Use company-cfo as a review lens for budget and spend questions |
| Content ops | Use jab-hook and watch-video patterns for content rotation across CorpusIQ social surfaces |
| Skill authoring | Use skillify and the meta-skills when converting recurring agent workflows into skills |
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.
Limitations / Verification¶
- Below the usual 20K publisher bar; drafted on suite-scale precedent (23 cohesive skills, 4.7K publisher-page installs, 681⭐, active development the day of the sweep) and direct relevance to business-operator agents
- The repo README and playbook prose target Claude Code, Codex, and Cursor; Hermes installs via the paths above, and tool references (e.g.
deep-researchnaming WebSearch, WebFetch, Notion) map onto equivalent Hermes tools - Snyk flags a Warn on the security audit (Gen Agent Trust Hub and Socket pass) - review SKILL.md files before activation
- Install counts are from skills.sh publisher pages as of the August 24, 2026 sweep; API search undercounts this cluster ~10x
# Verify skills installed
ls ~/.hermes/skills/makerskills/
# Verify the skills.sh listing
npx skills find "company-brain" --json 2>&1 | grep skills.sh
Security¶
- coreyhaines31/makerskills repo - review SKILL.md files before install (standard practice)
- Hermes skills security - skill trust guidance
- skills.sh security audits: Gen Agent Trust Hub Pass, Socket Pass, Snyk Warn
Related¶
- GBrain Agent Operations Setup - vector memory layer for Hermes agents
- Honcho Integration Setup - conversational memory infrastructure
- Agentic Awesome Skills (AAS) Setup - 2,000+ skill catalog with memory-systems playbooks
- Skills Catalog - full catalog index
← Skills Catalog | Marketplace →
Powered by CorpusIQ