Cli AnythingBLOCK
"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
CLI-Anything: Making ALL Software Agent-Native
  
Today's Software Serves Humans👨💻. Tomorrow's Users will be Agents🤖. CLI-Anything: Bridging the Gap Between AI Agents and the World's Software
🌐 [CLI-Hub](https://hkuds.github.io/CLI-Anything/): pip install cli-anything-hub then cli-hub install — browse, install, and manage all community-built CLIs. Want to add your own? Open a PR — the hub updates instantly.
🎬 See Demos: Watch AI agents use generated CLIs plus preview, live preview, and trajectory loops to produce real artifacts —
08db4c1acaf7OBSERVED · 2026-10-20Install
Commands as the repository documents them. They are shown, not run.
pip install git+https://github.com/HKUDS/CLI-Anything.git#subdirectory=3MF/agent-harness
pip install -e ".[dev]"
pip install git+https://github.com/HKUDS/CLI-Anything.git#subdirectory=3MF/agent-harness
git clone https://github.com/HKUDS/CLI-Anything.git
pip install -e .
pip install cli-anything-hub
Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned | |
| codex | mentioned | |
| copilot | mentioned | |
| cursor | mentioned | |
| gemini-cli | mentioned | |
| openclaw | mentioned | |
| windsurf | mentioned |
What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
--- name: cli-hub-meta-skill description: >- Discover agent-native CLIs for professional software. Access the live catalog to find tools for creative workflows, productivity, AI, and more. --- # CLI-Hub Meta-Skill CLI-Hub is a marketplace of agent-native command-line interfaces that make professional software accessible to AI agents. ## Quick Start ```bash # Install the CLI Hub package manager pip install cli-anything-hub # Browse all available CLIs cli-hub list # Search by category or keyword cli-hub search image cli-hub search "3d modeling" # Install a CLI cli-hub install gimp # Show details for a CLI cli-hub info gimp ``` ## Workflow Matrices A single CLI is one tool. A **matrix** is a whole workflow packaged as capabilities × providers — e.g. `video-creation` maps intents like `text.transcribe` or `visual.generate` to harness CLIs, public CLIs, Python libraries, native binaries, and cloud APIs. Reach for a matrix when a task spans several tools (produce a video, design an image, build a game). Standard agent sequence — **preflight before you install**: ```bash cli-hub matrix list # browse all matrices cli-hub can "transcribe audio" # find the capability across matrices cli-hub matrix search "video subtitle" # search; shows the matched capability cli-hub matrix preflight video-creation --json # what's usable here? (exit 3 = gaps) cli-hub matrix preflight video-creation -c text.transcribe --fix-hints # one capability + install hints cli-hub matrix install video-creation --capability text.transcribe # install ONLY what the task needs # After install, the matrix SKILL.md renders locally with provider-selection rules — read it. ``` Scope every install — do not bulk-install a 14-CLI matrix for a one-capability task. Use `--capability <id>`, `--recipe <id>`, or `--only a,b`, and `--dry-run` to preview the plan with zero side effects. `--json` is available on every matrix subcommand; exit codes are `0` ok · `3` partial/gaps · `1` failure · `2` usage error. Retry failures with `cli-hub matrix install <name> --resume`, and audit an install with `cli-hub matrix doctor <name>`. ## Live Catalog **URL**: [`https://reeceyang.sgp1.cdn.digitaloceanspaces.com/SKILL.md`](https://reeceyang.sgp1.cdn.digitaloceanspaces.com/SKILL.md) The catalog is auto-updated and provides: - Full list of available CLIs organized by category - One-line `cli-hub install` commands for each tool - Complete descriptions and usage patterns ## What Can You Do? CLI-Hub covers a broad range of software and codebases, empowering agents to conduct complex workflows via CLI: - **Creative workflows**: Image editing, 3D modeling, video production, audio processing, music notation - **Productivity tools**: Office suites, knowledge management, live streaming - **AI platforms**: Local LLMs, image generation, AI APIs, research assistants - **Communication**: Video conferencing and collaboration - **Development**: Diagramming, browser automation, network management - **Content generation**: AI-powered document and media creation Each CLI provides stateful operations, JSON output for agents, REPL mode, and integrates with real software backends. ## How It Works `cli-hub` is a lightweight wrapper around `pip`. When you run `cli-hub install gimp`, it installs a separate Python package (`cli-anything-gimp`) with its own CLI entry point (`cli-anything-gimp`). Each CLI is an independent pip package — `cli-hub` simply resolves names from the registry and tracks installs. ## How to Use 1. **Install cli-hub**: `pip install cli-anything-hub` 2. **Find your tool**: `cli-hub search <keyword>` or `cli-hub list -c <category>` 3. **Install**: `cli-hub install <name>` (installs the `cli-anything-<name>` pip package) 4. **Run**: `cli-anything-<name>` for REPL, or `cli-anything-<name> <command>` for one-shot 5. **JSON output**: All CLIs support `--json` flag for machine-readable output ## Example Workflow ```bash # Install the hub pip install cli-anything-hub # Find what you need cli-hub search video # Install it cli-hub install kdenlive # Use it with JSON output cli-anything-kdenlive --json project create --name my-project ``` ## More Info - Live Catalog: https://reeceyang.sgp1.cdn.digitaloceanspaces.com/SKILL.md - Web Hub: https://clianything.cc - Repository: https://github.com/HKUDS/CLI-Anything
Trust audit
BLOCKgrade F · trust 36/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | FAIL |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (4 observation(s))
- Network
- declared (6 observation(s))
- Shell
- declared (5 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (25)
cli-anything-iterm2 session send "python3 -m http.server 8000"
saved_metrics = pickle.load(f)
@click.option("--insecure", is_flag=True, help="Disable SSL verification")params=params, timeout=timeout, verify=False)
STS2_Bridge.dll
module = importlib.import_module(module_name)
__import__(lib)
__import__(pkg if pkg != "PIL" else "PIL.Image")
"""Helpers for safe JSON writes."""
- Install AdGuard Home: `curl -s -S -L https://raw.githubusercontent.com/AdguardTeam/AdGuardHome/master/scripts/install.sh | sh -s -- -v`
- Install AdGuard Home: `curl -s -S -L https://raw.githubusercontent.com/AdguardTeam/AdGuardHome/master/scripts/install.sh | sh -s -- -v`
Install AdGuard Home: `curl -s -S -L https://raw.githubusercontent.com/AdguardTeam/AdGuardHome/master/scripts/install.sh | sh -s -- -v`
twinkle_twinkle_G.mscz
twinkle_twinkle_G.mxl
exec(sync_script.read_text(encoding="utf-8"), namespace)
return hashlib.md5(email.strip().lower().encode()).hexdigest()
const REPO_ROOT = path.resolve(__dirname, "../../..");
tokens: '../../unsafe-tokens.json'
tokens: '../../../tokens/brand-test.json' // Correct relative path to reach sketch/agent-harness/tokens
"../../Uni-Mol/unimol_tools/unimol_tools/weights"
"../../../Uni-Mol/unimol_tools/unimol_tools/weights"
story="Dish pivot, radar plate, beacon core, and nav lights add the recognizable inspection payloads.",
label="Author hover, spin, and beacon motion",
story="Hover motion, dish spin, ring rotation, and beacon pulses prepare the final presentation state.",
"id": "beacon-cone",
Gates applied: critical_finding, no_behavioural_pass, undeclared_transfer.
08db4c1acaf7full audit observations/trust-audit/skill/hkuds__cli-anything.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-20 | 08db4c1acaf7 | BLOCK | F | 36 | first audit |
Questions
What does the cli-anything — hkuds/cli-anything skill do?
"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/
Is cli-anything — hkuds/cli-anything safe to install?
No — not without reading the findings first. The audit graded it F (36/100) and found 4 critical or high issues in the source. Each one is listed on this page with the file and line it is on.
What can cli-anything — hkuds/cli-anything access on my machine?
The audit observed that it reaches the network, runs shell commands and reads or writes files. Each of those is consistent with what it says it does. Secrets in the source: none found.
What do I need installed to use cli-anything — hkuds/cli-anything?
Its own instructions reference cli-anything-gimp. Dependencies are not all pinned to exact versions, so what installs today may differ tomorrow.
Which assistants does cli-anything — hkuds/cli-anything work with?
Its documentation mentions claude-code, codex, copilot, cursor, gemini-cli, openclaw and windsurf. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (08db4c1acaf7), read on 2026-10-20. The repository is watched, and a new audit runs when it changes — this is the first audit.
Add this badge to your README
Shows this page's audit result and links back to it. It updates when the audit does.
[](https://modelsatlas.com/skills/hkuds/cli-anything/)
Not your listing, or something out of date? Claim or update your listing.