The open hub for 130+ agent-ready CLIs. Desktop apps, SaaS APIs, Google Workspace and more — pip install any tool in seconds. Fork-and-extend the Hub to add your own.
Works with the tools you already use
Docker
GitHub
Discord
Telegram
Stripe
Notion
Drive
Gmail
Zoom
Shopify
Figma
Blender
Steam
TikTok
X
Vercel
Cloudflare
Jira
HubSpot
Spotify
YouTube
Airbnb
DoorDash
YelpText commands perfectly match LLM output format. Supports chained complex workflows, with JSON output giving Agents zero-cost parsing.
No extra dependencies, cross-platform. --help auto-documentation lets Agents discover capabilities independently — no manual introduction needed.
LibreOffice truly generates PDFs, Blender truly renders 3D, Inkscape truly exports SVGs. Zero compromise, zero toy implementations.
REPL interactive mode maintains session state; subcommand mode supports scripts and pipelines. Built-in undo/redo covers all scenarios.
Unit tests + real software E2E tests. 1,508 tests passing at 100%. Magic byte verification — not just a demo.
pip install -e . goes straight into PATH. cli-anything-gimp / blender / inkscape... ready out of the box.
Image editing: crop, filters, layer compositing, batch watermarks, format conversion. Uses real Pillow + GEGL backend.
3D modeling and rendering: scene management, object creation, material settings, CYCLES ray-trace rendering via bpy.
Vector graphics: SVG creation and editing, shapes/text/gradients/transforms, exports SVG/PNG/PDF via direct XML manipulation.
Digital painting: brush engine, layer management, HDR support, batch export via Krita Python scripting API.
Batch retouching, PSD layer manipulation, smart objects, export pipelines via Adobe UXP JavaScript API.
Vector art automation: artboard management, path operations, gradient fills, SVG/PDF export via UXP scripting.
Export frames/components, list design tokens, manage comments and versions via Figma REST API.
Template population, brand kit management, design export in bulk via Canva Connect API.
Live streaming and recording: scene/source management, recording control, stream config via JSON scene files + obs-websocket.
Audio processing: cut, mix, denoise, equalize, format conversion (MP3/WAV/FLAC) via Python wave + sox.
Video editing: multi-track timeline, filter chains, picture-in-picture compositing. Direct MLT XML manipulation with melt rendering.
Video editing: timeline editing, transition effects, subtitle overlays, color grading via MLT XML + melt renderer.
Color grading, timeline editing, Fusion VFX nodes, audio mixing via DaVinci Python scripting API.
Sequence assembly, effect automation, multicam editing, export presets via Adobe UXP ExtendScript.
Video transcoding, audio extraction, thumbnail generation, stream concatenation — high-level ffmpeg Python wrapper.
Playback control, playlist management, stream capture, format conversion via VLC HTTP interface + RC mode.
List, upload, download, share files; create folders; search — all via Drive API v3. Part of gworkspace suite.
Send, list, search, reply, trash emails; manage labels — Gmail API v1. +send +reply +triage helper shortcuts.
List/create/delete events; +agenda (upcoming N days); manage multiple calendars — Calendar API v3.
Read/write/append cell ranges, create spreadsheets, batch update — Sheets API v4. Perfect for data pipelines.
Create, read, append to Docs; extract plain text; get metadata — Docs API v1. Great for report generation.
List spaces, send and read messages — Chat API v1. Build internal notification bots with minimal setup.
Office suite: Writer docs, Calc spreadsheets, Impress presentations. Exports real PDFs with headless mode support.
Word / Excel / PowerPoint / Outlook — full COM automation: create, edit, convert to PDF, send mail, manage calendar.
Create/read/update pages, append blocks, query databases with filters, search workspace via Notion REST API.
Note creation, vault search, tag management, backlink graph queries via Obsidian Local REST API plugin.
Meeting management: create/list/delete meetings, manage participants, get recordings via Zoom REST API (OAuth2).
Post messages, upload files, list channels, manage reactions, search messages via Slack Web API (slack-sdk).
Send messages, manage guilds/channels/roles/members, create webhooks via Discord REST API v10.
Send text/photo/document/poll/location, manage groups, pin messages, set bot commands via Telegram Bot API (pure REST).
Post to channels, manage meetings, list members, create tabs via Microsoft Graph API.
Container lifecycle, image build/pull/push, volume management, log streaming via Docker Python SDK.
Create issues/PRs, trigger Actions workflows, manage releases, gists via GitHub REST API v3 (PyGithub).
Open files, diff views, install/list extensions, run tasks, format code via VS Code CLI + Extension API.
Screenshot, PDF print, JS eval, network intercept, form automation via Chrome DevTools Protocol (CDP).
Manage issues, MRs, pipelines, deployments, registry via GitLab REST API with project-scoped tokens.
Manage projects, run inspections, format code, install plugins via JetBrains Toolbox + IDE REST API.
Run/pull/delete models, stream chat completions, batch embeddings, manage local LLM library via REST API.
Chat completions, function calling, embeddings, image generation (DALL·E), audio transcription via OpenAI API.
Queue workflows, upload images, monitor generation, retrieve results via ComfyUI local REST API.
AI content generation: slides, documents, charts, websites in one click via AnyGen REST API (anygen.io).
Flowcharts and architecture diagrams: mxGraph XML operations, shapes/connectors/styles, exports PNG/SVG via draw.io CLI.
Upload/download files, manage folders, share with permissions, search by query via Google Drive API v3.
Upload/download, share links, manage folders, sync status via Microsoft Graph API with MSAL auth.
Upload/download, list folders, share links, manage team members via Dropbox Python SDK.
S3 bucket operations: upload/download, presigned URLs, lifecycle rules, cross-region replication via boto3.
Compress/extract archives, list contents, test integrity, split/join, encrypt with AES-256 via 7z subprocess wrapper.
Game library management, achievements, player stats, friends list, market prices via Steam Web API. Supports SteamCMD server deployment.
Browse game library, check free game offers, manage installations and account data via EGS OAuth2 REST API.
Asset upload, DataStore CRUD, Place publishing, user notifications via official Roblox Open Cloud API (rbxcloud).
LoL summoner/match/ranked data, Valorant stats, TFT leaderboards via official Riot REST API. Covers all Riot titles.
RCON server control: whitelist/ops, world commands, scheduled backups, player management. Supports Java & Bedrock editions.
Dedicated server control via RCON: map rotation, player management, config apply, stats. CS2 peaked 882K concurrent players.
WoW character data, Overwatch 2 stats, Diablo IV profile, Hearthstone collections via Blizzard official OAuth2 REST API.
Manage streams, create clips, query subscriptions, update channel info, poll/prediction management via Twitch Helix API.
Daily check-in, redeem codes, character/weapon stats for Genshin Impact, Honkai: Star Rail and Zenless Zone Zero via HoYoAPI.
Player stats, game history export, daily puzzle, tournament results via Chess.com Published-Data API (no key required).
Export games/tournaments, opening database queries, puzzles, broadcasts via Lichess open REST API. 100% free & open-source.
Query WR leaderboards, runner profiles, game categories, recent runs via Speedrun.com REST API (no auth for public data).
Check your game library, achievements, Game Pass catalog, friend activity and cloud gaming status via Xbox Live REST API. Never miss a free Game Pass drop again.
Browse DRM-free game library, compare versions, check GOG deals and freebies via GOG Games DB API + gogdl CLI. Your DRM-free paradise, CLI-controlled.
Player stats, shop rotation, Battle Pass progress, tournament standings via Epic internal API + fortnitepy. Check today's shop before dropping 1000 V-Bucks.
Playback control, playlist CRUD, track search, recommendations, Now Playing info via Spotify Web API. Let the agent build your perfect coding playlist.
Upload videos, manage metadata, bulk update captions, fetch analytics, handle comments via YouTube Data API v3. Automate your channel operations at scale.
Dedicated server control via RCON: spawn items, manage world, whitelist vikings, trigger world saves. Build your longship empire from the terminal.
Customers, payments, subscriptions, refunds, products, prices — full Stripe API via official Python SDK.
Products, orders, customers, inventory, discounts, fulfillment via Shopify Admin REST/GraphQL API.
SOQL queries, leads/contacts/opportunities CRUD, Apex execution, reports via Salesforce REST/Bulk API.
Contacts, deals, companies, tickets, email sequences, analytics via HubSpot CRM REST API.
Create/update issues, sprint management, JQL search, board queries, transitions via Jira REST API.
Ad campaigns, audiences, creatives, budget management, analytics via TikTok Business API.
Campaign/ad group/keyword CRUD, bid strategy, performance reports via Google Ads Python SDK (v29).
Post tweets, search, manage DMs, analyze engagement, schedule posts via X API v2 (OAuth 2.0).
SMS/MMS/WhatsApp messaging, voice calls, phone number management, call recordings via Twilio REST API.
Transactional email, template management, suppression lists, stats, inbound parsing via SendGrid REST API.
Lists/subscribers, campaigns, automations, reports, audience segmentation via Mailchimp Marketing API.
Deploy projects, manage domains, env vars, edge functions, analytics via Vercel REST API.
App metadata, review replies, release management, subscription reports via Google Play Developer API.
Ticket CRUD, user management, macros, views, satisfaction reports via Zendesk REST API.
DNS records, firewall rules, Workers deploy, cache purge, R2 storage, D1 databases via Cloudflare REST API.
Nearby store locator, menu lookup, nutritional data via McDonald's unofficial REST API. Order 20 Chicken McNuggets programmatically — I'm Lovin' It.
Store finder, menu items, combo deals, nutritional info via KFC regional APIs. Finger Lickin' Good terminal edition — let your agent handle the bucket list.
Food delivery, hotels, movie tickets, bike rental via 美团开放平台 API. China's super-app — let your agent order 外卖 at 3am without waking you up.
Restaurant search, delivery tracking, order management via Alibaba Ele.me open API. Your AI agent can argue with the delivery guy so you don't have to.
Store discovery, delivery estimates, order status, menu browsing via DoorDash Drive API. Late-night snack procurement, fully automated by your agent.
Restaurant search, menu items, delivery ETA, order tracking via Uber Eats Orders API. Because your agent deserves a meal break too.
Ride-hailing, GrabFood delivery, GrabPay transactions across SE Asia via Grab Open Platform API. SEA's everything-app, now CLI-native.
Ride booking, fare estimation, trip history via DiDi Open Platform API. Hail a cab with one CLI command — no surge pricing negotiation needed.
Property search, availability check, pricing calendar, host management via Airbnb unofficial API. Let the agent scout your next workation spot while you sleep.
Hotel/apartment search, availability, price comparison, review analysis via Booking.com Affiliate API. Score the best deal without opening 47 browser tabs.
Business search, reviews, ratings, location-based discovery via Yelp Fusion API. Find the best ramen near the office before the team lunch debate starts.
Scan source code, map GUI operations to underlying APIs, understand software architecture and capability boundaries.
Architect command groups, state models, JSON output formats — ensuring interface consistency and extensibility.
Build Click CLI with REPL sessions, JSON output, undo/redo — production-quality code.
Generate TEST.md covering unit tests + E2E test plans including real software invocation validation.
Implement full test suites with magic byte validation, ZIP structure verification, audio RMS detection — never trusting exit(0).
Update TEST.md with test results, generate app-specific SOP docs for Agents to self-discover capabilities.
Create setup.py, pip install -e . installs to PATH, cli-anything-[name] is ready to use.
# Use Python API to batch-generate reports → export PDF from cli_anything.inkscape.core import document, shapes, text, export proj = document.create_document("poster", width=1200, height=675) shapes.add_rect(proj, x=0, y=0, width=1200, height=675, style="fill:#0d0d1a") shapes.add_circle(proj, cx=900, cy=337, r=200, style="fill:#00ff41;opacity:0.1") text.add_text(proj, text="CLI-Anything", x=100, y=240, font_size=96, fill="#ffffff") document.save_document(proj, "poster.json") export.export_svg(proj, "poster.svg") from cli_anything.inkscape.utils.inkscape_backend import export_svg_to_png export_svg_to_png("poster.svg", "poster.png", width=1200, height=675) # → poster.png 89KB, 1200×675, rendered by real Inkscape
# REPL interactive mode — all changes share session state $ cli-anything-inkscape inkscape> document new --width 1200 --height 675 -o poster.json inkscape> shape add-rect --x 0 --y 0 -w 1200 -h 675 -s "fill:#0d0d1a" inkscape> text add -t "CLI-Anything" --x 100 --y 240 --font-size 96 --fill "#ffffff" inkscape> document save inkscape> export svg poster.svg --overwrite inkscape> exit # LibreOffice — generate PDF report $ cli-anything-libreoffice document new -o report.json --type writer $ cli-anything-libreoffice --project report.json writer add-heading -t "Q1 Report" --level 1 $ cli-anything-libreoffice --project report.json export render output.pdf -p pdf
# AI Agent integration: --json returns structured data import subprocess, json def agent_query(cmd): result = subprocess.run( ["cli-anything-inkscape", "--project", "design.json", "--json"] + cmd, capture_output=True, text=True ) return json.loads(result.stdout) # Agent reads current document state state = agent_query(["document", "info"]) # → {"name":"design","counts":{"objects":12,"layers":3},...} # Agent decides and executes if state["counts"]["objects"] < 5: agent_query(["shape", "add-circle", "--cx", "600", "--cy", "337", "--r", "100"]) # --help lets Agent self-discover all capabilities caps = subprocess.run(["cli-anything-inkscape", "--help"], capture_output=True, text=True).stdout
Isolate-test each core function with synthetic data, covering edge cases, parameter validation, and error paths.
Test project file generation pipelines: validate file format legality without invoking real software.
Invoke real software and validate output artifacts. Check magic bytes, pixel values, audio RMS.
Invoke installed commands via subprocess.run, validating JSON output format and return codes.
═════════════════════════════════ Test Summary ═════════════════════════════════ gimp 107 passed ✓ (64 unit + 43 e2e) blender 208 passed ✓ (150 unit + 58 e2e) inkscape 202 passed ✓ (148 unit + 54 e2e) audacity 161 passed ✓ (107 unit + 54 e2e) libreoffice 158 passed ✓ (89 unit + 69 e2e) obs-studio 153 passed ✓ (116 unit + 37 e2e) kdenlive 155 passed ✓ (111 unit + 44 e2e) shotcut 154 passed ✓ (110 unit + 44 e2e) zoom 22 passed ✓ (22 unit + 0 e2e) drawio 138 passed ✓ (116 unit + 22 e2e) anygen 50 passed ✓ (40 unit + 10 e2e) ─────────────────────────────────────────────────────────────────────────────── TOTAL 1,508 passed ✓ 100% pass rate 1,073 unit tests + 435 end-to-end tests
Tests FAIL (not SKIP) when backend is missing. Ensures CLI functionality is real — prevents false passes.
PDF validates %PDF- magic bytes; ZIP/OOXML verifies directory structure; PNG samples pixel values; audio detects RMS level.
Non-integer frame rates (29.97fps) accumulate errors — use round() not int(). Tests allow ±1 frame tolerance.
GUI software applies effects only on render. After CLI generates project file, must invoke real renderer — cannot substitute Pillow for GIMP.
Local codebase or GitHub URL
/cli-anything
./your-software
Auto-map APIs, understand architecture
Click CLI + REPL + JSON
Unit + E2E, full validation
pip install into PATH
Select Your AI Platform
# Step 1: Add CLI-Anything to plugin marketplace /plugin marketplace add chatjesus/CLI-Anything-Hub /plugin install cli-anything # Step 2: Generate CLI for any software (all 7 stages) /cli-anything:cli-anything ./my-software /cli-anything:cli-anything https://github.com/user/repo # Step 3: Iterative refinement (gap analysis + expand coverage) /cli-anything:refine ./my-software /cli-anything:refine ./my-software "batch processing and filters" # Step 4: Validate and test /cli-anything:test ./my-software /cli-anything:validate ./my-software
# Copy command files to OpenCode commands directory cp CLI-Anything/opencode-commands/*.md ~/.config/opencode/commands/ cp CLI-Anything/cli-anything-plugin/HARNESS.md ~/.config/opencode/commands/ /cli-anything ./my-software /cli-anything https://github.com/user/repo /cli-anything-refine ./my-software "picture-in-picture workflow" /cli-anything-list
# Install Codex Skill git clone https://github.com/chatjesus/CLI-Anything-Hub.git .\CLI-Anything\codex-skill\scripts\install.ps1 # Windows PowerShell # After restarting Codex, describe tasks in natural language: "Use CLI-Anything to build a harness for ./my-software" "Use CLI-Anything to refine ./shotcut for picture-in-picture"
# Register plugin git clone https://github.com/chatjesus/CLI-Anything-Hub.git bash CLI-Anything/qoder-plugin/setup-qodercli.sh /cli-anything:cli-anything ./my-software /cli-anything:refine ./my-software "batch processing" /cli-anything:validate ./my-software
Iterative Improvement: /refine Command
Automatically analyzes gaps between current CLI and software's full capabilities, identifies uncovered modules, incrementally implements new commands.
Specify a target domain for deep coverage — precisely add commands, tests, and docs for that scenario.
Each refine is incremental — never deletes existing commands. Run multiple times to progressively increase coverage.
New commands automatically get test cases added. TEST.md updates in sync, keeping docs and code consistent.
Types of Targets You Can Build
pip installschema command — no auth neededdetect command — check connectivity--json flag on all outputssetup.py with pip entry pointWe review, merge, and your CLI appears in the Hub — instantly usable by any AI agent via pip install.
schemaOutput full capability JSON. No auth needed. Agents call this to self-discover your CLI.
detectCheck connectivity without side effects. Returns JSON status object.
--jsonAll commands output structured JSON. Agents parse this programmatically.
versionPrint CLI version. Required for compatibility and update tracking.
No community CLIs yet — be the first contributor.
+ Submit Your CLI| Feature | GUI Screenshot | Limited REST API | CLI-Anything ✓ |
|---|---|---|---|
| Calls Real Software | ✓ | Partial | ✓ Full feature set |
| Stability | ✗ Pixel-clicks break easily | ✓ | ✓ Command-level stable |
| Structured Output | ✗ Requires OCR | ✓ JSON | ✓ Built-in --json |
| Agent Self-Discovery | ✗ | Requires reading docs | ✓ --help self-describing |
| Installation Cost | Medium | API Key | ✓ pip install |
| Feature Coverage | ~80% (UI-dependent) | ~20% (API limits) | ✓ Near 100% |
| Open Source / Commercial | Varies | Varies | ✓ MIT License |
$ git clone https://github.com/chatjesus/CLI-Anything-Hub.git $ cd CLI-Anything-Hub
$ winget install Inkscape.Inkscape # install desktop software $ cd inkscape/agent-harness $ pip install -e .
$ cli-anything-inkscape # enter REPL $ cli-anything-inkscape document new -o my.json # or CLI mode $ python -m pytest cli_anything/inkscape/tests/ -v
Fork CLI-Anything Hub to add your own tools, publish to PyPI, and grow the community. Fully open source under MIT.