Build for machines
first.
A registry for CLI tools, MCP servers, APIs, and SDKs designed to be operated by AI agents. Structured output. Predictable interfaces. Automated verification.
What makes a tool agent-friendly?
Eight core principles that transform tools from human-only interfaces into reliable building blocks for autonomous agents.
JSON Output by Default
Machine-readable output as the primary format. No parsing HTML tables or human-formatted text.
Structured Error Handling
Typed errors with specific exit codes. Agents can programmatically handle failures.
Natural Language Input
Parse human text into structured commands. Accept dates like "tomorrow" or "next friday".
Brief/Summary Commands
Quick context retrieval. Agents need fast overviews without fetching everything.
Bulk Operations
Handle multiple items in a single call. Reduce round-trips and latency.
Self-Contained
No external service dependencies. Local-first with portable data stores.
Health & Introspection
Let agents understand tool state. Doctor commands for self-diagnosis.
Deep Links
Stable, unique references to resources. Enable cross-tool coordination.
Discover agent-ready tools
CLI tools, MCP servers, APIs, and SDKs verified for agent compatibility. Ready to integrate into your autonomous workflows.
things-agent-cli
haasonsaas/things-agent-cli
Agent-first task management CLI inspired by Things. Local SQLite storage, natural language capture, recurring tasks.
filesystem
modelcontextprotocol/servers
MCP server for secure file system operations. Read, write, search, and manage files with controlled access.
browser-use
browser-use/browser-use
Make websites accessible for AI agents. Structured browser automation with accessibility tree parsing.
Test your tool's agent-readiness
Run our automated verification suite against your repository. Get instant feedback on which principles your tool follows.
Enter repo URL
Paste your GitHub repository URL or owner/repo format
Automated analysis
We scan your README, CLI help output, and code for patterns
Get your score
See which principles you follow and how to improve
Register your tool
Add your agent-friendly tool to the registry. We'll run automated verification and help agents discover your work.