claude-cli-runner

Execute Claude CLI prompts non-interactively with retry logic and file-based inputs.

14|6|Updated Feb 24, 2026
One-click install
npx skills add https://github.com/fkguo/nullius --skill claude-cli-runner
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude-cli-runner
Source: https://github.com/fkguo/nullius/tree/main/skills/claude-cli-runner
Command: npx skills add https://github.com/fkguo/nullius --skill claude-cli-runner

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

The Skill addresses the need to execute the Claude CLI in non-interactive mode for various prompts, providing functionality for retry logic, system and user prompts loading from files, and output redirection.

Core Features & Use Cases

  • Non-Interactive Mode: Run Claude CLI without manual interaction for automated tasks.
  • File Inputs: Load system and user prompts from files.
  • Output Redirection: Direct the output to a target file.
  • Retry Logic: Handle transient failures with exponential backoff.
  • Use Case: Utilize the skill for workflows that require interaction with Claude from a different host or environment, enabling cross-model work without incurring latency or context loss.

Quick Start

Execute the Claude CLI for a given prompt with retries using the skill, providing system and user prompts from files, and specifying the output file.

Frequently Asked Questions about claude-cli-runner

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

FAQPage Schema
How do I run the Claude CLI non-interactively for automated workflows?

You can run the Claude CLI non-interactively by using this skill to execute arbitrary prompts without manual interaction. It supports loading system and user prompts from files, directing output to a target file, and automating cross-model workflows.

Can I load system and user prompts from files when using the Claude CLI?

Yes, you can load system and user prompts from files when executing the Claude CLI. This skill supports file-based prompt loading, allowing you to manage complex prompt structures externally and redirect the output to a specified target file.

What is the best way to handle transient failures during non-interactive Claude CLI execution?

Handling transient failures during Claude CLI execution is done through built-in retry logic. This skill provides optional retries with exponential backoff, ensuring transient errors do not disrupt automated workflows without session persistence.

Do I need to install the Claude CLI separately to use non-interactive execution with retries?

Yes, you need to install the Claude CLI separately and configure it for session management and model selection. This skill requires the CLI environment to be set up, as it acts as a runner for non-interactive prompts without handling the installation.

Does non-interactive Claude CLI execution support session persistence across different hosts?

No, non-interactive Claude CLI execution does not support session persistence. This skill is designed for cross-model workflows from separate hosts or environments, meeting CLI interaction needs without retaining session state between executions.