Guide
By AnyCap Team
Install AnyCap
Fastest current path: ask your agent to help me install anycap.ai.
The fastest AnyCap setup flow today is natural language. In supported agents such as Claude Code or Cursor, you can ask the agent to install AnyCap and let it handle skill discovery, CLI setup, authentication, and verification. If you prefer direct command control, this page also includes the complete manual install path from CLI binary to first capability call.
Prerequisites
- A compatible coding agent if you want the natural-language install path
- macOS, Linux, or Windows (Git Bash)
- Node.js 18+ (if installing via npm or skills.sh)
- A browser for the login flow
Natural language is the fastest current setup path on supported agents. Manual CLI and skill commands remain below for direct control or unsupported environments.
Fastest path: ask your agent to install AnyCap
# Example prompt
help me install anycap.ai
On supported agents, this can trigger skill discovery, CLI install, authentication, verification, and your first capability call without memorizing commands. Start here if you use Claude Code or Cursor. If you want explicit terminal control, continue with the manual steps below.
Install the CLI manually
# Install script (recommended)
curl -fsSL https://anycap.ai/install.sh | sh
# Or via npm
npm install -g @anycap/cli
Binaries are also available on GitHub Releases.
Log in
# Login
anycap login
This opens a browser for authentication. After login, credentials are stored securely on your machine.
Install the skill file manually
# skills.sh (generic)
npx -y skills add anycap-ai/anycap -y
# skills.sh (agent-specific)
npx -y skills add anycap-ai/anycap -a claude-code -y npx -y skills add anycap-ai/anycap -a cursor -y npx -y skills add anycap-ai/anycap -a codex -y
# Or via AnyCap CLI
anycap skill install --target ~/.agents/skills/anycap-cli/
# Or manually
curl -fsSL https://raw.githubusercontent.com/anycap-ai/anycap/main/skills/anycap-cli/SKILL.md --create-dirs -o ~/.agents/skills/anycap-cli/SKILL.md
Skills teach AI agents how to discover and invoke AnyCap capabilities. Use the agent-specific skills add command when you know the host agent. Use the generic target path when you need manual control over where the skill file lives.
Verify
# Verify setup
anycap status npx -y skills check
This confirms CLI version, authentication state, server connection, and skill installation.
Use your first capability
# Or ask your agent
generate a product mockup image with AnyCap
# Understand an image
anycap image read --url https://example.com/photo.jpg
# Generate an image
anycap image generate --model seedream-5 --prompt "a sunset over mountains"
Common Issues
Natural-language install does not work in my agent
Use the manual CLI and skill steps on this page, or follow an agent-specific guide for Claude Code, Cursor, or Codex.
Install script fails on macOS
Try installing via npm instead: npm install -g @anycap/cli
Login does not open a browser
Use device code flow: anycap login --headless
Permission denied
Ensure the install directory is writable, or use npm with a user-level global prefix.
Common Questions
Can I install AnyCap with natural language?
Yes. In supported coding agents such as Claude Code and Cursor, you can say 'help me install anycap.ai' and let the agent handle skill discovery, CLI setup, authentication, and verification. If you prefer direct command control, the manual path is still available on this page.
Do I still need the AnyCap CLI if I use natural language?
Yes. Natural language is the control surface, but the AnyCap CLI is still the runtime that installs locally, stores authentication, and executes capabilities such as image generation, video generation, image understanding, and video analysis.
Which skill install command should I use?
Use npx -y skills add anycap-ai/anycap -a <agent> -y when you know the host agent, such as claude-code, cursor, or codex. Use the generic npx -y skills add anycap-ai/anycap -y when you want a broader default path, or use anycap skill install --target when you need manual control over the destination.