Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1,490 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

agent-ready-repo

The supervised agentic build loop—and the catalogue of workflows around it.

Decide what to build, design the product and system, implement and review software, plan infrastructure, validate releases, document what ships, and govern the catalogue your organization owns.

Reversible agent work → mechanical checks → independent review → human decision

PyPI License OWASP Agentic Skills Top 10

Build with core · Explore use cases · Get started · Browse the catalogue · Contribute

Choose a use case

  • Product managers and strategists: move from strategy and evidence to a bounded decision brief with product-strategy, desk-research, and product-engineering.
  • Platform, infrastructure, and SRE teams: design the boundary, author contracts, produce reviewable Terraform, and validate the deployed result with architect, contracts, iac-terraform, and release-engineering.
  • Software teams: take a brief or spec through implementation, hard gates, independent review, and a human merge decision with core.
  • Design, research, documentation, and enablement teams: explore the use cases that map your work to the relevant packs.

Packs are composable units, not departments. Install only what you need, or use a profile when a curated starting set fits. The complete catalogue remains the source for current pack inventory and install commands.

Build with core

core is the catalogue's flagship and its strongest reason to start. It is a complete build system for coding agents: spec and plan before code, mechanical gates before claims, a cold independent review before handoff, and a human decision before merge.

python -m pip install agentbundle
agentbundle install --pack core

Then ask your coding agent:

Implement <small, testable feature> with the work-loop.

The loop keeps fixing until the checks and review are clean. If it repeats the same failure, crosses a boundary, or reaches an irreversible decision, it stops and surfaces the choice instead of self-certifying. See how core compares with prompt-and-approve coding, Spec Kit, and Kiro.

Using Claude? You can also add this catalogue as a plugin marketplace and install any user-scope pack directly. Repo-scoped packs such as core install with agentbundle so the whole team shares the same loop.

Own the system

agentbundle projects portable skills, subagents, hooks, commands, and seeds into the layout each supported agent expects. Installed files remain readable and editable; upgrades preserve local changes as explicit companions instead of silently clobbering them.

Create an organization-owned catalogue without forking this repository:

agentbundle catalogue init my-catalogue --name my-catalogue
agentbundle catalogue verify --root my-catalogue

Understand packs, profiles, adapters, composition, and self-hosting · Create a catalogue · Read the authoring standards

Go deeper

License

Licensed under either Apache 2.0 or MIT at your option. Contributions are dual-licensed under the same terms unless you state otherwise.

About

The complete AI operating model for software teams — from first idea to production. Three peer-supervised loops (discovery → build → release) over a catalogue of curated packs: skills, subagents, and hooks, each installed in one line. It's npm for your coding agent. Any agent, any stack — Claude Code, Codex, Cursor, Copilot, Gemini, Kiro.

Topics

Resources

Contributing

Stars

17 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages