claw

Execute prompts against NinjaClaw agent containers from the terminal.

3|Updated Apr 19, 2026
One-click install
npx skills add https://github.com/OfirGavish/NinjaClaw-Nano --skill claw-ofirgavish
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claw
Source: https://github.com/OfirGavish/NinjaClaw-Nano/tree/main/.copilot/skills/claw
Command: npx skills add https://github.com/OfirGavish/NinjaClaw-Nano --skill claw-ofirgavish

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Claw provides a simple, command-line interface to send prompts directly to NinjaClaw agent containers from the terminal, enabling workflows without a chat app.

Core Features & Use Cases

  • Directly dispatch prompts to registered NinjaClaw groups by name, folder, or JID from the command line.
  • Auto-detects container or Docker runtime with optional overrides and easy PATH integration.
  • Read group configurations from the NinjaClaw database and load secrets from .env for secure, automated interactions.
  • Ideal for scripting, automation, and quick ad-hoc agent interactions in development or operations.

Quick Start

Install the claw script into your NinjaClaw directory, symlink it into your PATH, then run claw --list-groups to verify connectivity.

Frequently Asked Questions about claw

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run prompts against agent containers from the terminal?

To run prompts against agent containers from the terminal, you can use a CLI tool that pipes a JSON payload directly into a container run. This allows scripted interactions with registered agent groups by name, folder, or JID without needing a chat app.

How do I automate interactions with registered NinjaClaw groups?

You can automate interactions with registered NinjaClaw groups by executing a CLI script that reads group configurations from the database and loads secrets from a .env file. This securely pipes payloads into container runs for automated workflows.

Can I use Docker or other container runtimes to execute agent prompts?

Yes, you can execute agent prompts using Docker or other container runtimes. The CLI auto-detects the available container runtime and supports an optional --runtime override, allowing flexible execution across different environments.

Do I need a chat app to send prompts to NinjaClaw agents?

No, you do not need a chat app to send prompts to NinjaClaw agents. A command-line interface tool can dispatch prompts directly to agent containers, which is ideal for quick ad-hoc interactions and automated scripting.

What is the best way to manage secrets when scripting container agent interactions?

The best way to manage secrets when scripting container agent interactions is by loading them from a local .env file. The CLI automatically reads these environment variables to securely authenticate and run the container payloads.

How do I verify connectivity to registered agent groups from the command line?

To verify connectivity to registered agent groups from the command line, install the script into your directory, symlink it into your PATH, and run the list-groups command. This confirms the CLI can successfully read the agent database.