Skip to main content

Workloom

The package manager for AI workflows

Publish, version, and sync AI agent configurations across your entire organization. Every provider, every repo, one source of truth.

$ wl pull
✓ Resolved 3 packages for workspace "frontend-team" (revision 12)
✓ Rendering for targets: root, apps/web, apps/api
  → Claude: CLAUDE.md, .claude/settings.json, .claude/skills/review.md
  → Codex:  AGENTS.md, .codex/config.toml
  → Copilot: .github/copilot-instructions.md
✓ Wrote 6 files, updated .loom/state.json

$ wl sync --check
✓ All files match desired state (revision 12)
  Exit code: 0
4
Providers
Claude, Codex, Copilot, Gemini
11
Asset Types
Instructions, skills, agents, workflows, and more
1
Source of Truth
Your registry, every team
0
Config Drift
Enforced in CI

How It Works

1

Publish

Write canonical assets, publish versioned packages to your registry. Every configuration is immutable and auditable.

→
2

Assign

Assign packages to workspaces, configure targets for each repo. Control exactly what goes where.

→
3

Sync

Run wl pull to render the right files for every provider. CI checks enforce perfect consistency.

Before & After

Before Workloom

  • ✗Copy-paste configs across repos
  • ✗Manual onboarding for every new dev
  • ✗Silent drift with no detection
  • ✗No rollback when changes break
  • ✗Locked into one provider

With Workloom

  • ✓One publish + pull for all repos
  • ✓Instant onboarding with wl pull
  • ✓CI enforcement catches drift immediately
  • ✓Immutable versions with instant rollback
  • ✓Provider-agnostic configurations
📦

One Package, Every Provider

Write AI workflow configs once in a canonical format. Workloom renders them correctly for Claude, Codex, Copilot, and Gemini automatically.

🔒

Immutable Versions, Instant Rollback

Published package versions never change. Track every revision. Roll back in seconds with wl rollback. Full audit trail included.

🛡️

Drift Detection in CI

wl sync --check in your pipeline catches configuration drift before it reaches production. Zero manual verification needed.

🔐

Trust-Gated Security

Hooks, MCP servers, and tool permissions require explicit approval. Nothing that expands agent authority happens silently.

👥

Team-Wide Consistency

Publish your team's AI standards as a package. Every developer gets the same configuration on wl pull. Onboarding takes seconds.

🏗️

Monorepo Native

Multiple targets per workspace. Root, apps, and services each get their own provider configs. No cross-target overwrites.

Ready to stop managing AI configs by hand?

Join teams who ship faster with version-controlled AI workflows.