9 Best Claude Code CLI Tools to Pair With Your Agent in 2026

Claude Code is only as fast as the shell around it. The agent reads files, runs commands, and shells out to whatever you have installed. In 2026, the difference between a Claude Code session that ships and one that burns tokens is not the model. It is the CLI tools you pair with it.

The hard part is separating the tools that actually reduce tokens and round-trips from the marketing-flavored "AI dev tools" that just wrap a chat box. We have run Claude Code daily across client engagements, internal repos, and the AY Automate platform, and we have watched which CLIs consistently cut session time, reduce context bloat, and prevent the agent from doing dumb things.

This guide compares the 9 best Claude Code CLI tools in 2026. Real workflows, honest pricing where it is publicly known, pros and cons, and a framework to pick the right combination for your repo.

Best Claude Code CLI tools: a brief overview

• ripgrep (rg): Best for fast code search inside the agent loop: the de facto replacement for grep, used by Claude Code internally and worth installing globally.
• fzf: Best for interactive file and command selection: a fuzzy finder that pairs with rg, git, and your shell history to feed Claude Code precise paths.
• gh CLI: Best for GitHub-native agent workflows: lets Claude Code open PRs, read issues, and triage CI without leaving the terminal.
• repomix: Best for one-shot repo-to-prompt packing: bundles a repo into a single XML/Markdown file ready to paste into any model or feed to Claude Code.
• files-to-prompt: Best for surgical context selection: Simon Willison's tool that flattens a chosen file set into one prompt-ready blob.
• code2prompt: Best for token-budgeted context: Rust CLI that builds prompts from a codebase with token counts, templates, and .gitignore respect.
• bun: Best for fast JS/TS execution: an all-in-one runtime, installer, and bundler that lets Claude Code iterate Node projects without npm slowness.
• uv: Best for Python project setup at agent speed: an extremely fast Python package and project manager from Astral that replaces pip + venv + pyenv.
• rtk (Rust Token Killer): Best for cutting Claude Code token spend: a CLI proxy that filters noisy command output before it reaches the agent.

1. ripgrep (rg), best for fast code search inside the agent loop

ripgrep is the search engine Claude Code actually wants. The agent shells out constantly to find symbols, references, and patterns. With plain grep, every search is slow and noisy. With rg, it is fast, respects .gitignore, and returns clean output that the agent can parse in a single pass.

If you only install one tool from this list, install ripgrep. It is also a hard prerequisite for many other agent workflows, including most "code RAG" setups and the Claude Code internal file tools.

Key features

• Recursive search faster than grep, ag, and ack on most workloads
• Respects .gitignore, .ignore, and .rgignore by default
• Unicode and PCRE2 regex support, JSON output for scripting
• First-class glob filtering: rg "TODO" -g "*.ts" -g "!**/dist/**"
• Single static binary, ships in every major package manager

Best for

• Any Claude Code user on a non-trivial repo
• Teams that want consistent search behavior across macOS, Linux, and Windows
• Anyone running code-search-heavy agent loops (refactors, audits, migrations)

Pricing

• Free and open source (MIT)
• Available via brew, apt, cargo, winget, and direct binaries

Pros

• Massive speedup over default tools
• Zero configuration to be useful
• Output format is agent-friendly: stable, parseable, deterministic

Cons

• No interactive UI on its own; pair with fzf for picking results
• Regex syntax differs slightly from grep/PCRE, which trips up some scripts

2. fzf, best for interactive file and command selection

fzf is a fuzzy finder that turns any list into a searchable picker. For Claude Code, it is less about the agent calling fzf directly and more about you piping the agent precise paths and commands without manually typing them.

Combined with ripgrep, git, and shell history, fzf becomes the connective tissue of an agent-friendly terminal. Pick a file in one keystroke, hand it to Claude Code, get a focused edit.

Key features

• Fuzzy matching with score-based ranking and preview pane
• Integrates with shell history (Ctrl-R), file picker (Ctrl-T), and cd (Alt-C)
• Pipe-friendly: works with anything that emits lines (git log, rg, find)
• Supports multi-select, custom keybinds, and preview windows

Best for

• Engineers who want a faster terminal feedback loop around Claude Code
• Teams that already use ripgrep and want a cohesive search-pick-edit workflow
• Anyone tired of manually copy-pasting paths from ls output

Pricing

• Free and open source (MIT)
• Install via brew install fzf, apt install fzf, or cargo install

Pros

• Tiny binary, zero config, runs everywhere
• Composes with every CLI you already have
• Preview pane shrinks the gap between "find" and "open"

Cons

• The agent itself does not "use" fzf; the benefit is on the human side of the loop
• Default keybindings overlap with some shell setups and need tweaking

3. gh CLI, best for GitHub-native agent workflows

gh is GitHub's official CLI and the cleanest way to let Claude Code work with PRs, issues, releases, and Actions. Instead of pasting URLs and screenshots, you let the agent run gh pr view, gh pr create, gh run watch, and stitch the GitHub side of the workflow directly into the build loop.

For teams shipping daily, gh is the missing piece that turns Claude Code from "writes code locally" into "owns the whole PR".

Key features

• Full coverage of issues, PRs, releases, Actions, gists, and Codespaces
• JSON output (--json) for clean parsing by agents
• Auth via OAuth or PAT, with org and SSO support
• Extensions ecosystem (gh extension install) for custom workflows

Best for

• Any team on GitHub running Claude Code in real engineering work
• Solo developers using the AY Framework /ship and /review modes
• Engineering teams that want PR descriptions and changelog entries drafted by the agent

Pricing

• Free and open source
• No GitHub plan changes required

Pros

• Officially maintained by GitHub
• JSON mode is precise enough for safe agent automation
• Eliminates context switches between terminal and browser

Cons

• Some advanced GraphQL queries still require gh api, which the agent can mishandle
• Rate limits matter once you have the agent polling CI status loops

If you already use Claude Code for shipping PRs, our Claude Code agency team's standard setup pairs gh with the /ship workflow defined in our best Claude Code workflows guide.

4. repomix, best for one-shot repo-to-prompt packing

repomix bundles an entire repo (or a filtered subset) into a single file optimized for LLM consumption. It is the cleanest way to take a codebase, strip the noise, count tokens, and hand the result to a model that needs the full picture.

For Claude Code users, repomix is most useful when you want to feed a sibling model (Claude.ai web, ChatGPT, a local Llama) the same context, or when you need a permanent snapshot of a codebase for a planning session.

Key features

• One command produces XML, Markdown, or plain-text output of an entire repo
• Built-in token counting via tiktoken
• File exclusion patterns, .gitignore respect, and binary detection
• Security checks for secrets via Secretlint integration
• Direct integration with Claude, ChatGPT, and Gemini web flows

Best for

• Pre-planning sessions where you want context outside Claude Code
• Audits where the auditor uses a different model than the implementer
• Open-source maintainers who want a "share this repo with an LLM" link

Pricing

• Free and open source (MIT)
• Run via npx repomix or install globally with npm

Pros

• Sensible defaults out of the box, no config required
• Token counting prevents accidentally blowing past context limits
• Works on any repo, any language

Cons

• Single-shot output: not designed for incremental, in-loop use like Claude Code's own tools
• Large monorepos still need manual filtering to stay under token budgets

5. files-to-prompt, best for surgical context selection

Simon Willison's files-to-prompt is the minimalist alternative to repomix. You point it at specific files and directories and it concatenates them into a single prompt-shaped blob, with optional Claude XML formatting.

Where repomix tries to be comprehensive, files-to-prompt is precise. When you know exactly which 8 files matter, this tool gives you the cleanest payload with zero ceremony.

Key features

• Accepts files, directories, and globs; respects .gitignore
• Output formats: plain, Markdown, and Claude-style XML (--cxml)
• Optional line numbers (-n) for precise reference
• Pipe-friendly: chains with rg, find, and shell substitution

Best for

• Power users who already know which files matter
• Quick one-off prompts in Claude.ai web that need 5-20 file payload
• Scripted agent workflows that need deterministic context

Pricing

• Free and open source (Apache 2.0)
• pip install files-to-prompt or uvx files-to-prompt

Pros

• Tiny, single-purpose, easy to script
• Claude XML output works perfectly with Claude.ai web
• Pairs naturally with find, rg, and fzf

Cons

• No automatic token counting (use ttok or claude-tokenizer-cli separately)
• Less polished than repomix for whole-repo dumps

6. code2prompt, best for token-budgeted context

code2prompt is a Rust CLI that builds prompts from a codebase with first-class token counting, Handlebars templates, and .gitignore support. It sits between repomix and files-to-prompt: more programmable than either, with better budgeting for tight context windows.

It is the right tool when you need consistent prompt structure across many runs (audits, refactors, codemods) and want to enforce a token budget per prompt.

Key features

• Token counting with tiktoken for GPT and Claude models
• Handlebars template engine for custom prompt structure
• Glob include/exclude, .gitignore and .code2promptignore support
• Output to stdout, file, or clipboard
• Single static Rust binary

Best for

• Teams running repeated audits or migrations at scale
• Engineers building custom agent pipelines around Claude or GPT
• Anyone hitting context window limits with naive prompt building

Pricing

• Free and open source (MIT)
• Install via cargo install code2prompt or pre-built binaries

Pros

• Templating is genuinely useful for repeatable workflows
• Token visibility prevents silent truncation
• Fast: Rust binary, no Node or Python runtime

Cons

• Templates have a small learning curve compared to repomix defaults
• Less integrated with web UIs than repomix

7. bun, best for fast JS/TS execution

bun is an all-in-one JavaScript runtime, package manager, bundler, and test runner. For Claude Code working in Node, TypeScript, or React projects, swapping npm install and node script.ts for bun install and bun script.ts removes a huge chunk of agent wait time.

Slow installs and slow test runs are not just human-painful: they cost real money in agent sessions, because the agent sits idle waiting for output. bun fixes that.

Key features

• Drop-in Node-compatible runtime, often 2-5x faster on real workloads
• Built-in package manager with binary lockfile and parallel installs
• Native TypeScript and JSX execution: no transpile step
• Built-in test runner, bundler, and bunx package runner
• First-class Web APIs (fetch, WebSocket, FormData)

Best for

• TypeScript and Node projects where install/test loops dominate session time
• Side-project repos where simplicity beats convention
• Edge and serverless code that targets Bun, Cloudflare Workers, or Deno

Pricing

• Free and open source (MIT)
• curl -fsSL https://bun.sh/install | bash or brew install oven-sh/bun/bun

Pros

• Genuine speed gains, especially on installs and cold starts
• Reduces agent idle time during loops
• Single binary replaces node, npm, tsx, ts-node, jest

Cons

• Some Node-native modules and tooling still have rough edges
• Production parity with Node still requires verification per project

For TypeScript-heavy agent projects, our Claude Code agency team standardizes on bun for the dev loop and Node for production.

8. uv, best for Python project setup at agent speed

uv is Astral's Python package and project manager, written in Rust. It replaces pip, venv, pyenv, pip-tools, and poetry with one fast binary. For Claude Code working in Python repos, uv removes minutes of latency per session.

If you run any Python automation, RAG pipeline, or data scripts under Claude Code, uv is the single biggest upgrade you can make to the shell environment.

Key features

• 10-100x faster than pip on real workloads
• Built-in virtualenv management (uv venv)
• Project management with uv init, uv add, uv sync
• Python version management (uv python install 3.13)
• uvx for ephemeral tool execution (replaces pipx)
• Single static Rust binary

Best for

• Any Python developer running Claude Code on real projects
• RAG and agent framework work (LangChain, LlamaIndex, CrewAI)
• Data and ML engineers who want fast environment bootstrapping

Pricing

• Free and open source (Apache 2.0 / MIT)
• curl -LsSf https://astral.sh/uv/install.sh | sh or brew install uv

Pros

• Speed that is genuinely noticeable, not marginal
• Compatible with pyproject.toml and existing Python tooling
• Replaces 4-5 tools with one binary

Cons

• Project workflow (uv sync) is newer and still evolving
• Some legacy projects with custom setup.py flows need manual tweaks

9. rtk (Rust Token Killer), best for cutting Claude Code token spend

rtk is a CLI proxy that sits between Claude Code and the commands it runs. It rewrites and filters noisy command output (long npm install logs, verbose test runners, build chatter) before the agent ever sees it, cutting input tokens by 60-90% on common dev operations.

In long Claude Code sessions, the dominant cost is not the model's thinking. It is the agent re-reading bloated tool output. rtk attacks exactly that.

Key features

• Transparent hook: Claude Code calls commands normally, rtk filters before return
• Per-command rules: trim install logs, dedupe test output, collapse progress bars
• rtk gain analytics: shows tokens saved per session and per command
• rtk discover mode: analyzes your history to surface missed savings
• Works with any tool: git, npm, pytest, docker, pnpm

Best for

• Power users running long Claude Code sessions (>1 hour)
• Teams paying real money for Anthropic API tokens
• Anyone tired of seeing 5,000 lines of npm output eat their context window

Pricing

• Free and open source
• Install via cargo install rtk or pre-built binaries

Pros

• Real, measurable token savings (the rtk gain command shows the math)
• Zero workflow changes for the agent
• Useful even outside Claude Code (reduces terminal scroll fatigue)

Cons

• Setup requires a Claude Code hook configured in settings.json
• Aggressive filtering can hide errors if rules are not tuned

How to choose the best Claude Code CLI tools

1) What is the minimum baseline every Claude Code user should install?

If you only install three tools, install ripgrep, gh CLI, and bun or uv (depending on your stack). ripgrep accelerates every search the agent runs, gh CLI gives the agent ownership of the GitHub side of your work, and bun/uv eliminates the install and test latency that dominates session cost. After that, add fzf for the human side of the loop and rtk once your sessions get long enough to feel the token bill.

2) When should I use repomix vs files-to-prompt vs code2prompt?

Use repomix for "give me the whole repo as one prompt" — quick exploration, audits, sharing context with another LLM. Use files-to-prompt when you know exactly which files matter and want a clean, scriptable payload (it pairs beautifully with ripgrep and find). Use code2prompt when you need repeatable prompt templates with token budgeting across many runs, like a batch refactor or an internal codemod pipeline. For ongoing in-loop work, Claude Code's own tools beat all three.

3) How do these tools cut Claude Code token spend?

The biggest token sinks in Claude Code are noisy command output, repeated file reads, and oversized context payloads. rtk directly attacks output noise. ripgrep keeps searches scoped (no node_modules reads). bun and uv finish commands faster, so the agent reruns less. code2prompt and files-to-prompt enforce token budgets when you do step outside Claude Code. Combined, these tools typically cut session token usage by 40-70% on real engineering work. For deeper patterns, see our roundup of best Claude Code skills.

4) Do I need all nine, or is this overkill?

No team needs all nine on day one. Start with ripgrep + gh CLI + your language-specific speed upgrade (bun or uv). Add fzf when manual file selection starts to slow you down. Add repomix or files-to-prompt the first time you need to feed a non-Claude-Code model. Add rtk once you have a real token bill to worry about. The point is not maximalism: it is matching the tool to the specific friction in your loop.

Ready to ship faster with Claude Code?

The right CLI stack turns Claude Code from a chat-with-files agent into a real engineering teammate. At AY Automate, we run this exact stack on every client engagement, tuned to the repo and the language. If you want a team that has already battle-tested these tools on production codebases, our Claude Code agency builds, ships, and maintains Claude Code workflows end to end. Pair this guide with our best Claude Code GitHub repos for templates and our best Claude Code skills for the prompt-level upgrades. When you are ready to talk, book a consultation and we will audit your current setup in 30 minutes.

FAQ

What are Claude Code CLI tools?

Claude Code CLI tools are command-line utilities you install in your shell that Claude Code can call as part of its tool use loop, or that you call manually to feed Claude Code cleaner inputs. They include search tools (ripgrep), package managers (bun, uv), GitHub clients (gh), and context-shaping tools (repomix, files-to-prompt, code2prompt).

How are CLI tools different from Claude Code plugins or MCP servers?

CLI tools are generic command-line programs that any agent or human can call. Claude Code plugins extend Claude Code's behavior at the IDE/agent level. MCP servers expose structured tools over the Model Context Protocol. They are complementary: CLI tools sit in the shell, plugins sit inside Claude Code, and MCP servers sit between Claude Code and external systems. See our best Claude Code MCP servers for the MCP layer.

How do I verify a CLI tool is safe to run inside an agent loop?

Three checks. First, the tool should be open source and widely adopted (ripgrep, gh, uv, bun all have very large user bases). Second, it should have predictable output: deterministic, parseable, no interactive prompts. Third, you should be able to dry-run it: tools like repomix and code2prompt let you preview output before piping it anywhere.

How much do Claude Code CLI tools cost in 2026?

Every tool in this list is free and open source. The only cost is the Anthropic API or Claude Pro subscription that powers Claude Code itself. The whole point of the right CLI stack is to reduce that bill, not add to it.

How long does it take to set up a full Claude Code CLI stack?

Less than 30 minutes for the baseline (ripgrep + gh + bun or uv + fzf). Add repomix or files-to-prompt in another 5 minutes. rtk takes 10-15 minutes because it needs a Claude Code hook. Total: under an hour for a full setup that will pay for itself in the first long session.

Is the official Anthropic Claude Code release the same as Claude Agent SDK?

Claude Code is the IDE-grade agent CLI from Anthropic. The Claude Agent SDK is a lower-level library for building your own agents on the same primitives. CLI tools in this list work with both, but most of the day-to-day developer experience benefits land in Claude Code.

Should I use Claude Code CLI tools or just rely on Claude Code's built-in tools?

Use both. Claude Code's built-in tools (Read, Edit, Grep, Bash) cover the common path. The CLI tools in this list make that common path faster, cheaper, and more accurate. ripgrep makes the built-in Grep faster. bun and uv make the built-in Bash loop faster. rtk makes every command cheaper. They stack with Claude Code, not against it.

Can a Claude Code agency set this up for my team?

Yes. A good Claude Code agency will audit your current CLI stack, install and configure the right tools per repo, write the Claude Code hooks and skills that wire them together, and train your engineers on the new loop. At AY Automate, that is exactly what our Claude Code agency team does on every engagement. Book a consultation if you want to walk through your current setup.