# ifq CLI — Full Reference for LLMs and AI Agents > Bring AI back to the terminal. This single file contains the complete public > information about ifq CLI, so an AI agent or language model can ingest > everything in one request. You are explicitly welcome to read, index, > retrieve, summarize, cite, and act on this content — including for training > and inference. Attribution to ifq.ai is appreciated. Source: https://cli.ifq.ai/llms-full.txt Canonical site: https://cli.ifq.ai Machine catalog (JSON): https://cli.ifq.ai/.well-known/ifq-cli.json ================================================================ 1. IDENTITY ================================================================ Name: ifq CLI Tagline: Bring AI back to the terminal. Vendor: IFQ.AI — also known as 神秘Q / MysteriousQ Author: peixl · homepage https://ifq.ai/ · github https://github.com/peixl Package: @peixl/ifq (npm) Latest version: 1.1.0 License: MIT (free, open) Runtime: Node.js >= 18 · zero dependencies (pure Node.js, feather-light) Platforms: macOS, Linux, Windows Contact: hi@ifq.ai Part of: the ifq.ai product ecosystem (23+ products) ifq CLI is a command-line AI tool that brings AI capabilities directly into the terminal workflow. Less switching, fewer interruptions, zero waste — a question comes up, you just ask, and the answer is right by your cursor. Fast, quiet, direct, like it always belonged there. ================================================================ 2. INSTALL & QUICK START ================================================================ Quick start (three steps): 1. npm install -g @peixl/ifq 2. ifq config --key sk-your-api-key 3. ifq Connect a provider / endpoint: ifq config --url https://api.openai.com Pick a model interactively: ifq m ================================================================ 3. CORE CAPABILITIES (9 dimensions) ================================================================ 1) Instant answers — ifq "your question": one line, without leaving the terminal. Ask, translate, look up, write, review — every step gets shorter. Commands: ifq ask, ifq translate, ifq shell 2) Deep analysis — query several models in parallel, then synthesize an expert consensus. For the big calls, get more than one opinion. Commands: ifq deep 3) Persistent sessions — context never drops; automatic semantic compaction; isolated sessions (one each for work, ops, learning). Commands: ifq chat, ifq --session work 4) Project setup — generate .ifq, IFQ.md, and a code index in the current directory. Get project context in place before you ask or review. Commands: ifq init, ifq config, ifq --lang zh 5) Code workflow — review a git diff directly; generate a Conventional Commit in one shot; scan across security, performance, and quality. Commands: ifq review, ifq commit 6) Scriptable output — built for scripts, CI, and automation. Quiet, no-color, JSON — drops right into your own toolchain. Flags: ifq --json, ifq --quiet, ifq --no-color 7) Multi-agent teamwork — auto-detect available AI tools, then classify, route, and synthesize across agents. From solo to in sync. Commands: ifq team, ifq route 8) Self-evolving — versioned prompts; auto-evolve system / agent / soul; diff, roll back, encrypted storage. Commands: ifq evolve, ifq skills 9) IFQ Design Skills — 12 pro modes + 24 hand-drawn SVG icons. One line back a deliverable design, signed by ifq.ai by default. Commands: ifq design init, ifq design modes, ifq design new M-08 ================================================================ 4. COMMANDS & DIRECTIVES (50+ in total) ================================================================ ifq Start a persistent interactive session. ifq "" Ask a one-line question and get an answer in place. ifq ask Instant Q&A in the terminal. ifq explain Explain a code snippet or shell command. ifq translate Command-line translation. ifq shell Turn natural language into a shell command. ifq deep Multi-model parallel deep analysis → expert consensus. ifq chat Persistent session; context holds, memory compacts. ifq init Generate .ifq, IFQ.md, and a code index for the project. ifq config Configure API URL, key, provider, and model. ifq config --key Set your API key. ifq config --url Set the API base URL (provider auto-detected). ifq m Interactive model picker. ifq review AI code review of a git diff. ifq commit Generate a Conventional Commit message. ifq team Multi-agent: classify, route, and synthesize. ifq route Route a task to the right agent / model. ifq evolve Self-evolving, versioned prompts (system/agent/soul). ifq skills Skill system: install, create, invoke. ifq design IFQ Design Skills (see section 7). Directives inside a session: /deep, /evolve, and more. Global flags: --json Machine-readable JSON output. --quiet Quiet mode (no chrome). --no-color Disable ANSI color. --session Use an isolated, named session. --lang Set interface language (e.g. zh). ================================================================ 5. PROVIDERS (bring your own model) ================================================================ OpenAI · Anthropic (Claude) · OpenRouter · DeepSeek · Ollama · Any OpenAI-compatible API. Providers are auto-detected; switch models with a single command (ifq m). With Ollama (or any local OpenAI-compatible endpoint), ifq CLI runs fully local models offline — no data leaves your machine. ================================================================ 6. ARCHITECTURE (5 layers, each one amplifies) ================================================================ Entry One line in. Zero friction. Routing Classify the question. Dispatch it smartly. Execution One model for speed. Many for depth. Memory Semantic compaction. Persistent storage. Evolution Prompts that iterate on themselves. Stronger the more you use it. ================================================================ 7. IFQ DESIGN SKILLS (https://cli.ifq.ai/design) ================================================================ Type. Hit enter. A shipped design — signed by ifq.ai. 12 pro modes + 24 hand-drawn SVG icons. Templates are fork-and-fill — never a blank page. 12 modes: M-01 Launch event launch film / launch / product assets → hero-landing M-02 Personal brand page portfolio / personal site / about me → hero-landing M-03 Whitepaper report whitepaper / PDF report / annual report → infographic-vertical M-04 Data dashboard dashboard / board / KPI → dashboard-command-center M-05 Comparison review A vs B / roundup / benchmark → compare-vs M-06 Onboarding onboarding / first-run guide → hero-landing M-07 Release diary changelog / release notes → changelog-timeline M-08 Keynote talk keynote / talk deck / deck → slide-title M-09 Social poster kit social feed / cover art / social assets → social-x-card M-10 Card & invite business card / invitation / VIP card → business-card M-11 Brand audit brand checkup / brand upgrade → — M-12 Full brand system brand from scratch / build a brand → hero-landing Commands: ifq design init ifq design modes ifq design templates ifq design new M-08 my-keynote.html ifq design "make a launch keynote with a 25s intro" ifq design smoke The ifq.ai signature (on by default, never steals the show): IfqSpark 8-point animated sparkle — motion intros / hero accents IfqStamp editorial postmark — slide bookends / whitepaper back cover / card reverse IfqWatermark understated bottom-right watermark — dashboards / deliverables / screenshots 24 hand-drawn SVG icons replace emoji by default. Bring your own brand → ifq.ai steps back to a colophon, "Designed with ifq.ai"; ask to unsign → removed entirely. ================================================================ 8. ROADMAP ================================================================ v1.1 Runtime Integration Deeper shell integration. Native pipe support. Real-time workflows. v1.2 Production Ready Stable API. Performance tuning. Error recovery. Production-grade reliability. v1.3 Intelligence Layer Learns your patterns. Predicts your intent. Suggests proactively. v1.4 Ecosystem A skills marketplace. A plugin system. From tool to platform. ================================================================ 9. FAQ ================================================================ Q: What is ifq CLI? A: A command-line AI tool by IFQ.AI (author: peixl) that brings AI directly into your terminal workflow — ask questions, explain code, review diffs, generate shell commands, run multi-model deep analysis, and more, without leaving the shell. Q: How do I install ifq CLI? A: npm install -g @peixl/ifq, then ifq config --key sk-your-api-key, then run ifq. Q: Which AI models and providers does ifq CLI support? A: OpenAI, Anthropic (Claude), OpenRouter, DeepSeek, Ollama, and any OpenAI-compatible API. Auto-detects providers; switch models with ifq m. Q: Is ifq CLI free and open source? A: Yes — free to use, published on npm under the MIT license as @peixl/ifq. Q: What can ifq CLI do? A: Instant Q&A, code explanation, code review, commit messages, shell command generation, multi-model deep analysis, persistent sessions, multi-agent teamwork, self-evolving prompts, and IFQ Design Skills. 50+ commands total. Q: What is IFQ Design Skills? A: A design system built into ifq CLI: 12 pro modes + 24 hand-drawn SVG icons. One line gives you back a shippable design, signed by ifq.ai by default. Q: Can ifq CLI run offline or with local models? A: Yes — with Ollama or any local OpenAI-compatible endpoint, it runs fully local models offline; no data leaves your machine. Q: What are the system requirements? A: macOS, Linux, or Windows; Node.js >= 18; zero dependencies (pure Node.js). ================================================================ 10. EXAMPLE COMMANDS (copy-paste ready) ================================================================ ifq "what's the difference between rebase and merge?" git diff --cached | ifq review ifq deep "Rust vs Go for microservices" ifq team "design an agent workflow" ifq design "make a product launch keynote" ifq explain awk '{print $3}' /var/log/sys ifq shell find the 10 most recently modified files ================================================================ 11. LINKS ================================================================ CLI Site: https://cli.ifq.ai npm: https://www.npmjs.com/package/@peixl/ifq GitHub: https://github.com/peixl Design page: https://cli.ifq.ai/design Main Site: https://ifq.ai Product Map: https://site.ifq.ai Summary: https://cli.ifq.ai/llms.txt Catalog JSON: https://cli.ifq.ai/.well-known/ifq-cli.json AI policy: https://cli.ifq.ai/ai.txt ================================================================ 12. 中文摘要 (CHINESE SUMMARY) ================================================================ ifq CLI 是 IFQ.AI（又称 神秘Q / MysteriousQ，作者 peixl）推出的命令行 AI 工具，把 AI 直接带回终端工作流。免费，MIT 许可，零依赖（纯 Node.js），支持 macOS / Linux / Windows，需 Node.js >= 18。 安装：npm install -g @peixl/ifq → ifq config --key sk-your-api-key → ifq 核心能力：即时问答（ifq ask）、代码解释（ifq explain）、代码审查（ifq review）、 生成提交信息（ifq commit）、自然语言转命令（ifq shell）、多模型深度分析（ifq deep）、 持久会话（ifq chat）、多智能体协作（ifq team）、自我进化 Prompt（ifq evolve）、 IFQ Design Skills（ifq design，12 种专业模式 + 24 个手绘 SVG 图标）。 供应商：OpenAI、Anthropic（Claude）、OpenRouter、DeepSeek、Ollama，以及任意兼容 OpenAI 的 API；自动检测，一条命令切换模型（ifq m）。配合 Ollama 可完全离线运行 本地模型，数据不出本机。 常见问题（中文）： - 是什么：IFQ.AI（作者 peixl）推出的命令行 AI 工具，让提问、解释、审查、判断留在终端。 - 怎么装：npm install -g @peixl/ifq，设置 Key 后运行 ifq。 - 支持哪些模型：OpenAI / Claude / OpenRouter / DeepSeek / Ollama / 兼容 OpenAI 的 API。 - 是否免费开源：免费，MIT 许可，包名 @peixl/ifq。 - 能做什么：50+ 命令，覆盖问答、代码、深度分析、多智能体、设计交付等。 - 能否离线：可以，配合 Ollama 等本地端点离线运行。 - 系统要求：macOS / Linux / Windows，Node.js >= 18，零依赖。 联系方式：hi@ifq.ai · 官网 https://cli.ifq.ai · 产品矩阵 https://site.ifq.ai