Frequently Asked Questions

General

What is Kage?

Kage is a local-first agentic work orchestrator that enables Claude Code agents to work autonomously, share context, and scale across multiple repositories.

Why is it called Kage?

“Kage” (影) means “shadow” in Japanese. The name reflects the philosophy of shadow agents working invisibly, executing with precision while you focus on what matters.

Is Kage open source?

Yes, Kage is open source under the Apache 2.0 license.

Technical

What agents does Kage support?

Currently, Kage has first-class support for Claude Code. The trait-based architecture allows extending to other providers (OpenAI, Gemini, etc.) in the future.

Does Kage require a database?

No. Kage uses embedded storage (redb) that’s compiled into the binary. There are no external dependencies.

Does Kage work offline?

Yes. Kage is local-first and doesn’t require any cloud services to operate. However, your AI agents may need internet access to function.

Where is my data stored?

All data is stored locally on your machine:

• State: ~/.local/share/kage/
• Config: ~/.config/kage/
• Secrets: Your OS keychain (macOS Keychain, Linux Secret Service, Windows Credential Manager)

Security

How are secrets stored?

Secrets are stored in your operating system’s native keychain:

• macOS: Keychain
• Linux: Secret Service (GNOME Keyring, KWallet)
• Windows: Credential Manager

Secrets are never written to disk in plaintext.

Is it safe to run agents autonomously?

Kage provides guardrails like iteration limits and approval workflows. You can configure:

• Maximum iterations before pause
• Approval requirements for writes or commits
• Checkpoint intervals for review