Skip to main content
skill CX Codex

cli-power-tools — Claude Code skill: what it does, how to install

cli-power-tools guides coding agents to use fast Rust-based command-line utilities and structured parsers instead of legacy shell pipelines.

MIT 121 installs ★ 29 updated 2026-08-17

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add oakoss/agent-skills --skill cli-power-tools -a claude-code
Open on GitHub

What cli-power-tools does

This skill teaches agents to pick purpose-built terminal programs instead of writing brittle shell pipelines. Code searching relies on ripgrep to respect .gitignore rules out of the box. File discovery delegates to fd, file inspection uses bat with syntax highlights, and folder structures are inspected using eza.

Rather than forcing plain text through ad-hoc parsers, the skill channels queries into structured tools. It processes JSON records via jq, handles YAML configs through yq, builds table queries with Nushell, and handles HTTP calls using xh.

The skill ships only guidelines and invocation patterns. Every referenced binary must already be installed and discoverable in the system PATH for the agent commands to work.

What it helps with

  • search large codebases with ripgrep while honoring ignore files
  • execute batch text replacements across projects with fd and sed
  • query and transform JSON or YAML configurations using jq and yq
  • extract structured log events using rg --json piped directly into jq

Example prompts for the agent

  • Find all .tsx files referencing useAuth and rank them by match counts
  • List files modified within the last hour and preview them with bat
  • Convert config.yml into JSON format and filter down to the active database entries

How to install cli-power-tools

1Skills CLI (Vercel)

npx skills add oakoss/agent-skills --skill cli-power-tools -a claude-code

2Manually: put the files in place

~/.claude/skills/cli-power-tools/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/cli-power-tools/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about cli-power-tools

How to install cli-power-tools in Claude Code?

Run npx skills add oakoss/agent-skills --skill cli-power-tools -a claude-code. The skill definition will be placed in ~/.claude/skills/cli-power-tools/SKILL.md.

What does cli-power-tools do?

It instructs coding agents to employ modern CLI replacements such as ripgrep, fd, bat, eza, and jq instead of relying on standard grep, find, or raw awk scripts.

Does this skill install the CLI binaries automatically?

No, it only provides operational rules and command patterns. Binaries like rg, fd, bat, yq, and xh must be pre-installed on your system.

Source text

SKILL.md · MIT
---
name: cli-power-tools
description: 'Performs advanced CLI operations, structured data transformation, and Unix forensics with modern Rust-powered tools. Use when searching codebases with ripgrep, transforming JSON or YAML with jq, refactoring across files with sed and fd, debugging APIs with httpie, or navigating directories with zoxide.'
license: MIT
metadata:
  author: oakoss
  version: '1.0'
  source: 'https://github.com/BurntSushi/ripgrep'
user-invocable: false
---

# Utility Pro

## Overview

Masters the command-line environment by turning raw text and unstructured data into actionable insights. Combines Rust-powered search tools (ripgrep, fd, bat), structured data shells (Nushell), JSON manipulation (jq/yq), and modern HTTP clients (xh) for high-performance CLI workflows.

**Core philosophy:** Replace legacy text-parsing pipelines (grep/awk/sed chains) with purpose-built tools that understand structure. ripgrep understands gitignore rules, fd understands file types, jq understands JSON, and Nushell understands tables.

**When to use:** Codebase-wide search and analysis, multi-file refactoring, JSON/YAML transformation pipelines, API debugging, directory navigation, log parsing, disk usage auditing, dependency tracing.

**When NOT to use:** GUI-based tasks, complex application logic (write a script instead), tasks requiring interactive prompts.

## Quick Reference

| Pattern                  | Tool / Command                                                           | Key Points                                         |
| ------------------------ | ------------------------------------------------------------------------ | -------------------------------------------------- |
| Text search              | `rg "pattern" -g "*.tsx"`                                                | 10-100x faster than grep, respects `.gitignore`    |
| File search              | `fd -e pdf` or `fd "pattern"`                                            | Fast alternative to `find` with smart defaults     |
| Syntax-highlighted read  | `bat file.ts`                                                            | Cat with syntax highlighting and git integration   |
| Directory jump           | `z project-name`                                                         | Remembers frequently visited directories           |
| Fuzzy search             | `fzf` or `CTRL-T` / `CTRL-R`                                             | Interactive selection for files and history        |
| Tree view                | `eza --tree --level=2`                                                   | Metadata-rich ls replacement with tree support     |
| JSON query               | `jq '.items[] \| select(.active)'`                                       | Full functional programming for JSON               |
| YAML query               | `yq '.config.database' file.yml`                                         | jq-like syntax for YAML files                      |
| HTTP request             | `xh POST api.example.com/v1/data key=value`                              | Colorized, user-friendly HTTP client               |
| Multi-file refactor      | `fd -e tsx -x sed -i 's/Old/New/g' {}`                                   | Find files then execute transforms                 |
| Structured shell         | `ls \| where size > 1mb \| sort-by size`                                 | Nushell tables instead of text parsing             |
| Log parsing              | `rg "ERROR" --json \| jq 'select(.type == "match") \| .data.lines.text'` | Combine ripgrep JSON output with jq filters        |
| File preview with search | `fzf --preview 'bat --color=always {}'`                                  | Browse files with syntax-highlighted preview       |
| Find recently changed    | `fd --changed-within 1h`                                                 | Find files modified within a time window           |
| Export count by file     | `rg -c "pattern" -g "*.ts" \| sort -t: -k2 -rn`                          | Count matches per file, sorted by frequency        |
| JSON format conversion   | `yq -o json config.yml`                                                  | Convert between YAML, JSON, CSV, and other formats |
| Advanced regex search    | `rg -P "(?<=function\s)\w+" --only-matching`                             | PCRE2 for lookahead/lookbehind patterns            |
| Git-aware file listing   | `eza -la --git --icons`                                                  | Show file metadata with inline git status          |

## Tool Overview

| Category         | Modern Tool    | Replaces              | Key Advantage                            |
| ---------------- | -------------- | --------------------- | ---------------------------------------- |
| Text search      | ripgrep (`rg`) | grep, ag, git grep    | Speed, `.gitignore` awareness, Unicode   |
| File search      | fd             | find                  | Smart defaults, regex, parallelism       |
| File viewer      | bat            | cat, less             | Syntax highlighting, git diff markers    |
| Directory jump   | zoxide (`z`)   | cd, autojump          | Frecency-based ranking, fzf fallback     |
| Fuzzy finder     | fzf            | manual piping         | Universal selector for any list          |
| File listing     | eza            | ls, tree              | Tree view, git status, icons             |
| JSON processing  | jq             | awk, python oneliners | Typed, functional JSON transforms        |
| YAML processing  | yq             | sed on YAML           | jq-compatible syntax for YAML            |
| HTTP client      | xh             | curl, httpie          | HTTPie syntax, Rust speed, single binary |
| Structured shell | Nushell        | bash + awk/sed        | Typed tables, native format conversion   |

## Common Mistakes

| Mistake                                               | Correct Pattern                                                                    |
| ----------------------------------------------------- | ---------------------------------------------------------------------------------- |
| Using `grep` instead of `ripgrep` for codebase search | Use `rg` which is 10-100x faster and respects `.gitignore` by default              |
| Piping `ls` output into `grep` for file filtering     | Use `fd` for file discovery, or `eza -D` / `eza --ignore-glob` for listing         |
| Writing complex `awk` scripts for structured data     | Use Nushell or `jq` which natively understand JSON, CSV, and YAML                  |
| Running `rm -rf` in scripts without a dry-run step    | Always add a verification or dry-run step before destructive operations            |
| Using capturing groups in regex when not needed       | Prefer non-capturing groups `(?:...)` for better performance in large-scale scans  |
| Searching `node_modules` or `.git` directories        | Use `rg` which skips these by default, or configure exclusions explicitly          |
| Forgetting `select(.type == "match")` on `rg --json`  | ripgrep JSON emits begin, match, context, end, and summary message types           |
| Using `curl` with verbose flag syntax                 | Use `xh` which provides colorized output and HTTPie-compatible request-item syntax |
| Parsing text output between tools with `cut`/`tr`     | Use Nushell pipelines that preserve structured data through every stage            |
| Anchoring regex with `.*` at the start                | Anchor with `^` when possible; unanchored `.*` causes expensive backtracking       |

## Installation

All tools are available via common package managers. Most are single statically linked binaries.

| Tool    | macOS (Homebrew)       | Linux (apt/cargo)                                |
| ------- | ---------------------- | ------------------------------------------------ |
| ripgrep | `brew install ripgrep` | `apt install ripgrep` or `cargo install ripgrep` |
| fd      | `brew install fd`      | `apt install fd-find` or `cargo install fd-find` |
| bat     | `brew install bat`     | `apt install bat` or `cargo install bat`         |
| zoxide  | `brew install zoxide`  | `cargo install zoxide`                           |
| fzf     | `brew install fzf`     | `apt install fzf`                                |
| eza     | `brew install eza`     | `cargo install eza`                              |
| jq      | `brew install jq`      | `apt install jq`                                 |
| yq      | `brew install yq`      | `snap install yq` or `go install`                |
| xh      | `brew install xh`      | `cargo install xh`                               |
| Nushell | `brew install nushell` | `cargo install nu`                               |

Shell integration (add to shell config after installing):

| Tool   | Shell Integration Required                                |
| ------ | --------------------------------------------------------- |
| zoxide | `eval "$(zoxide init zsh)"` (or bash/fish equivalent)     |
| fzf    | Source keybinding and completion scripts from fzf install |

## Delegation

- **Large-scale codebase search and analysis**: Use `Explore` agent to run ripgrep queries, trace dependencies, and map code patterns
- **Multi-file refactoring workflows**: Use `Task` agent to coordinate fd, sed, and verification steps across an entire project
- **Pipeline architecture for data transformation**: Use `Plan` agent to design structured pipelines combining jq, Nushell, and API tools

## References

- [Modern Unix Toolbox](references/modern-unix-toolbox.md) -- ripgrep, fd, bat, zoxide, fzf, eza, terminal setup
- [Advanced Regex and jq](references/advanced-regex-and-jq.md) -- named captures, lookahead/lookbehind, jq filters, integrated pipelines
- [Nushell Structured Data](references/nushell-structured-data.md) -- table paradigm, API interaction, scripting, Unix interop

© oakoss · MIT · oakoss/agent-skills

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set106

The whole set →
usability-tester usability-tester plans usability test sessions, evaluates interfaces against heuristic standards, and scores UX friction points using severity formulas. CX Codex +1 122installs react-email Build responsive HTML emails from React Email components and render them as HTML or plain text. Useful for transactional messages and marketing templates. CX Codex +1 122installs rag-implementer rag-implementer guides an agent through a retrieval-augmented generation build, from chunking and embeddings to hybrid retrieval and quality metrics. CX Codex +1 121installs service-worker service-worker guides the coding assistant in configuring Workbox v7.4 caching, offline app shells, background sync, and push notifications. CX Codex +1 121installs quality-auditor quality-auditor runs a weighted audit across 12 dimensions, scores each one against cited evidence and hands back a prioritized fix list. CX Codex +1 120installs package-publishing package-publishing guides npm package metadata, ESM and CJS builds, and TypeScript declarations. Useful when preparing a library’s public entry points or publishing workflow. CX Codex +1 119installs svelte A reference sheet that teaches a coding agent Svelte 5 runes and SvelteKit routing, plus a table for moving Svelte 4 code onto the new API. CX Codex +1 117installs react-native Guides agents in building React Native and Expo apps for iOS, Android, and web using file-based routing, EAS workflows, and native device modules. CX Codex +1 117installs

Project sponsors

Companies that keep this analytics open