An agent catalog can look useful right up until the agent enters your
repository.
The card says “research agent.” The install button says “ready.” What I
actually need to know is less exciting:
Which tools can it call?
A practical test for reusable agents: copy the definition into your project, inspect every permission boundary, and accept that updates become your responsibility.
AgentsKit Registry promotes 'copy-the-source' agents: code is copied directly into projects as local modules, not installed from remote catalogs. Tech leaders gain complete code review of security—tool permissions, model provider, confirmation enforcement—and prevent remote updates from changing behavior without visibility.
An agent catalog can look useful right up until the agent enters your
repository.
The card says “research agent.” The install button says “ready.” What I
actually need to know is less exciting:
Which tools can it call?

If you’ve used Claude Code, Cursor, Codex, Aider, Gemini CLI, GitHub Copilot, Grok, goose, or similar...

Anthropic's Stainless acquisition and recent Claude Code updates show where AI coding is headed: safer tool access, scoped…

Introduction Everyone is talking about autonomous AI agents. Give an agent internet access, connect...

Previous parts of Build a Basic AI Agent From Scratch: Basic Agent Tools Long Task Planning Human...

The four tools referenced below are open source at github.com/officialwhitebird. When you run AI...

A desktop AI agent that writes its own tools — and why every line passes a gate before it runs. ...