Agent readiness signals
Score each CLI by installability, non-interactive behavior, structured output, auth model, safety boundaries, and docs quality.
Agent-ready CLI directory
Find CLI tools your AI agent can actually use
Search, evaluate, install, and teach Claude Code, Codex CLI, Gemini CLI, Copilot CLI, and humans to use command-line tools with predictable output and safety boundaries.
$ clifinder search "deploy nextjs" --agent codex --json
> found: vercel-cli, cloudflare-wrangler, github-cli
$ clifinder skill github-cli --format skill.md
Structured output, auth notes, risk flags, and copyable agent instructions.
20CLI tools
10categories
27Playbooks
2026-05-14Last verified
Why CLI Finder
A directory built for agent workflows, not passive browsing
Every tool card is shaped around the details a developer or coding agent needs before running commands.
Copyable operating notes
Turn tool pages into short instructions agents can follow in AGENTS.md or SKILL.md without reading a full manual first.
Use-case navigation
Start from tasks like deploy Next.js, test webhooks, scrape a site, manage PRs, query data, or inspect production logs.
CLI, MCP, and API context
Understand when a shell command is enough, when an MCP server is safer, and when both belong in the same workflow.
recommended / parseable / safer
Agent-ready picks
agent playbooks
Playbooks
guide Best CLI tools for AI agents
A practical starter stack of CLI tools that agents can install, parse, and use with approval boundaries.
guide What makes a CLI agent-ready?The practical rubric behind CLI Finder's Agent Readiness Score.
guide MCP vs CLI for AI agentsA decision matrix for choosing CLI, MCP, API, or a combined tool interface for agents.
agent Best CLI tools for Claude CodeA CLI stack, approval model, and AGENTS.md starter for Claude Code users.
comparison Claude Code vs Codex CLI for terminal workflowsCompare how two coding agents should be equipped with CLI tools.
use-case CLI tools for deploying Next.js appsA preview-first Next.js deployment workflow for humans and AI agents.
User feedback
Built for people who hand terminal work to agents
CLI Finder helps me decide which commands are safe to give a coding agent before it touches a repo.
The structured-output notes are the part I needed. They save my team from inventing tool rules from scratch.
It is more useful than a list of logos because it explains install paths, auth, and destructive commands in one place.
FAQ
Common questions
Is CLI Finder only for AI agents?
No. Human developers can use it as a practical CLI directory, but the evaluation criteria are optimized for workflows where an agent has to install, parse, and use commands safely.
Does the site run commands for me?
No. CLI Finder is a public directory and guide layer. It does not log in to your services, execute commands, or mutate your projects.
How should I use the scores?
Treat the Agent Readiness Score as a starting signal, then read the install, output, auth, and safety notes before adding a CLI to your agent rules.
Can a CLI author submit a tool?
The current MVP is a curated public directory. Submission, verification, and publisher workflows are outside the current product scope.
Developer tool categories
Developer Workflow
Tools for repositories, pull requests, issues, releases, and local development loops.
DeploymentCLI tools for preview deploys, production releases, environment sync, and edge platforms.
PaymentsTools for testing webhooks, billing events, payment APIs, and local integration loops.
AI CodingCommand-line agents and coding assistants for repository work, reviews, and local automation.
Data ProcessingTools for JSON, CSV, Parquet, logs, and repeatable command-line data transforms.
Web ScrapingTools for approved web extraction, markdown conversion, crawling, and source-backed research.
Start with a safer stack
Give your agent a smaller, clearer CLI toolbox
Browse recommended tools, then use the playbooks to decide which commands need review before an agent can run them.