Equipt

Equip your AI with 600+ verified skills & agents.

Equipt is an open-source, curated & verified marketplace of AI skills & agents for founders and operators — the work of running a business: marketing, sales, ops, finance, content. 501 skills + 106 agents across 5 plugins, each machine-scored on the Equipt Standard before it ships, so you add vetted tools to claude.ai (or Claude Code) instead of a dump of prompts.

→ Browse the catalog: equipt-agent.vercel.app

Install

On claude.ai / desktop (no code — the way most operators use it):

• A whole category at once: download a plugin zip (e.g. equipt-marketing.zip) from the latest Release, then Customize → Personal plugins (+) → Create plugin → Upload plugin and select it — every skill in the plugin goes live. (Paid plans. The plugin's agents install too — they activate in Cowork & Claude Code.)
• Just one skill: grab that skill's <name>.zip and use Customize → Skills (+) → Create skill → Upload a skill instead (works on Free too).

In Claude Code — via the plugin marketplace:

/plugin marketplace add Salah-XD/equipt
/plugin install equipt-marketing

With the Equipt CLI (for the dev-savvy) — drops skills & agents into a project's .claude/:

npx @equipt/cli init                   # scaffold .claude/ + a manifest
npx @equipt/cli add equipt-marketing   # a whole plugin — or: add cold-outreach-writer

The five plugins

• equipt-business — Operator skills and agents: finance, operations, HR, legal, client work, e-commerce, events.
• equipt-content — Writing skills and agents: blog posts, copywriting, email, scripts, courses.
• equipt-data — Data & analytics skills and agents: SQL, dashboards, analytics narrative.
• equipt-engineering — Subagents for developers: code review, debugging, testing, refactoring, security audits, API & schema design.
• equipt-marketing — Marketing skills and agents: ads, SEO, social, email, sales funnels, growth, branding.

Browse everything in the live catalog or docs/directory.md.

The Equipt Standard

Every skill & agent earns a Readiness score (0–100) across five axes — Craft · Fit · Guard · Proof · Upkeep — capped by a safety gate, and falls into one of three tiers:

Tier	Meaning
Provisional	Machine-scored — not yet human-reviewed
Certified	Human-reviewed · Readiness ≥ 70
Field-Ready	Top tier · Readiness ≥ 85 with strong Guard & Proof

Read the methodology at equipt-agent.vercel.app/standard.

Develop

npm install
npm test            # marketplace + scorer unit tests
npm run lint        # validate skill/agent frontmatter
npm run build       # regenerate manifests, directory, and this README's plugin list
npm run score       # score every asset on the Equipt Standard → scores/
npm run new-skill -- --plugin equipt-marketing --name my-skill --description "..."

Folders under plugins/ are the single source of truth; marketplace.json, each plugin.json, and docs/directory.md are generated — never edit them by hand. The CLI lives in cli/, the scorer in scoring/, the site in site/.

Contributing

Open a pull request to add or improve a skill/agent, or an issue to propose one — see Publish to Equipt.

License

Code: MIT · Skill & agent content: CC-BY-4.0.