Skip to content

Enterprise Quickstart

SkillMeat's enterprise edition has ten separate guides under guides/enterprise-* and guides/federation-*, each covering one capability in depth. None of them tells you which order to read them in. This page is that order: a funnel through the existing guides, sequenced for someone moving from local edition into an enterprise deployment for the first time.

This page does not restate what those guides say — each step below is a short "why you're here" orientation and a link to the guide that actually covers it. If you already know which guide you need, skip straight to it; use this page when you don't.

Prerequisites

Before starting the path below, confirm:

  1. Your organization has deployed enterprise edition. Enterprise requires SKILLMEAT_EDITION=enterprise and a PostgreSQL database — it is not something you configure from the CLI alone. If you're the one deploying it, see Enterprise Deployment and Server Setup first; this quickstart picks up after that's done.
  2. You have the SkillMeat CLI installed. If not, follow Installation in the local quickstart — the CLI is edition-agnostic.
  3. You have enterprise credentials. Either a Personal Access Token (PAT) from your admin, or an SSO (Clerk) account. Authenticate with:
skillmeat auth login --enterprise <url>

This prompts for a PAT and stores it in ~/.skillmeat/enterprise.toml. Run skillmeat auth login --help for the full option reference.

The Path

1. Orient yourself: what enterprise adds

Start here if this is your first time on an enterprise instance. This guide covers the five things enterprise adds over local edition (multi-team support, authentication, RBAC, governance, federation), how to switch teams, and the difference between your personal and team collections.

→ Using SkillMeat Enterprise Features

2. Get your first artifact onto the instance

Once you know your role and team, the next natural step is uploading something. This guide covers skillmeat enterprise add — the single-artifact upload command with create-or-update semantics (idempotent re-uploads, new version on content change).

→ Enterprise Add: Uploading Artifacts to Enterprise Instances

3. Import artifacts at scale, from GitHub or marketplace

If you're bringing in more than a handful of hand-authored files — an existing GitHub repo of skills, or a marketplace source — this guide covers the two import paths: client-relay (fetch locally, relay to the server; available now) and server-side fetch (server fetches directly; gated feature, admin-enabled).

→ Enterprise Live-Source Ingestion

4. Update artifacts without losing history

Once you have artifacts in the instance, the next question is how to change them safely. Deleting and re-adding an artifact destroys its version history and orphans deployment links — this guide covers the three safe update paths instead (metadata-only, single-file content, and bulk re-publish with conflict=overwrite).

→ Enterprise Artifact Safe Updates

5. Understand centrally-managed artifacts and enforcement

Some artifacts on your instance won't be owned by you or your team — they're owned by the enterprise itself, created by system admins, and may be locked against local modification. This guide covers what the "Enterprise Managed" badge and lock icon mean, who can create these artifacts, and how global deployment works.

→ Enterprise Governance Guide

6. Run the admin workflows: inspect, bulk-act, organize

If you administer a team or the whole organization, this is the operational guide: inspecting an artifact's governance/propagation/audit tabs, bulk actions across multiple artifacts, managing collections, and triaging the activity feed. It also covers the post-deployment seed validation check — run this once, right after your org's first enterprise deploy.

→ Enterprise Admin Workflow Guide

7. Manage who can do what

Once you're comfortable with day-to-day admin workflows, role administration is the next layer: changing a user's role (org_member / org_admin / enterprise_admin), deactivating accounts, and reading the audit log for role changes. Requires the admin:users scope.

→ Enterprise RBAC Role Administration

8. Monitor sync health and divergence across teams

With users, roles, and artifacts in place, the ongoing job is keeping versions consistent. This guide covers the Team Sync Dashboard and Admin Sync Management views — how to read divergence indicators, cross-axis divergence, and cascade policies across org → team → user tiers.

→ Enterprise Unified Sync UX

9. Promote and force-sync versions

When the sync dashboard flags divergence, this guide covers what to actually do about it: the promotion workflow (request → review → approve/reject), force-sync for immediate overrides, and bulk admin version actions.

→ Enterprise Version Management

10. Migrate deployment tokens off the legacy scope

Last, because it's a maintenance task rather than a core concept: if your organization uses federation deploy-pull operations, PATs using the legacy materialization:write scope need to move to project:deploy before the deprecation window closes. This guide covers identifying legacy tokens, minting replacements, and updating CI/CD.

→ Federation PAT Migration Guide

Not Yet Documented

None of the ten guides above cover bootstrapping the very first system_admin account or the first team on a freshly deployed enterprise instance. Using SkillMeat Enterprise Features explicitly says "you don't need to set up enterprise features yourself — your admin provisions teams" and Enterprise RBAC Role Administration confirms system_admin is "service-account-only... provisioned via seed tooling" — but neither says who runs that seed tooling or how, for a brand-new deployment with no admin yet. If you're standing up a fresh instance and hit this gap, check with whoever ran the deployment (see Enterprise Deployment) rather than expecting a guide to walk you through it.

Verify It Worked

A few concrete checkpoints as you move through the path above:

# Confirm you're authenticated and to which instance
skillmeat auth status

# If you're using named instances instead of enterprise.toml directly
skillmeat instance show

# Confirm your first upload landed (Step 2) — re-running is safe and idempotent
skillmeat enterprise add ./my-skill.md
# Second run should print something like:
#   = artifact skill:my-skill already up to date (no changes)

If you're an admin who just deployed the instance, also run the seed validation check described in Step 6 before doing anything else — admin operations (teams, roles, deployments) don't work correctly without it.

Reference

These aren't part of the reading path above — keep them open in another tab while you work through it.