# AgentBundle

> Define an AI agent once; ship it to all 7 supported runtimes — Claude Code, Cursor, GitHub Copilot, OpenCode, Gemini, Codex, Windsurf — with built-in security scans on every publish.

AgentBundle is a platform for defining AI agents once and shipping them across every team in your org. Anyone can author — engineers from the CLI, marketers and ops from the dashboard. Every publish runs through built-in secret and prompt-injection scanners.

Use `/llms-full.txt` for the full text dump. Every page is also available at `<path>/index.md` as raw markdown.

## Core
- [Pricing](https://www.agentbundle.dev/pricing): Tiers and limits
- [Use cases](https://www.agentbundle.dev/use-cases): Author → Review → Distribute → Use → Iterate narratives across roles
- [Security](https://www.agentbundle.dev/security): Built-in scans on every publish; honest compliance posture

## Writing
- [Blog](https://www.agentbundle.dev/blog): Engineering and product writing
- [Built-in security scans on every publish](https://www.agentbundle.dev/blog/2026-04-15-built-in-scans): Secret scanning and prompt-injection detection run on every agent publish, on every plan. Here is why that is the default.
- [Defining agents once, running them everywhere](https://www.agentbundle.dev/blog/2026-05-01-defining-agents-once): Why we built AgentBundle to define an agent in one place and ship it to every IDE — Claude, Cursor, Copilot, OpenCode.

## Reference
- [Contact](https://www.agentbundle.dev/contact): How to reach us

## Optional
- [Changelog](https://www.agentbundle.dev/changelog): AgentBundle product releases