skills.rest
aspenkit
aspenkitOFFICIAL
Explore all skills by this author →
WebsiteGitHub

claude-runner

Executes Claude/Codex prompts via CLI with stream-json output parsing and skill file generation.

95|4|Updated Feb 27, 2026
Search Tags:#cli#codex#tooling#claude#skill-generation#stream-json#prompt-loading
One-click install
npx skills add https://github.com/aspenkit/aspens --skill claude-runner
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: claude-runner
Source: https://github.com/aspenkit/aspens/tree/main/.agents/skills/claude-runner
Command: npx skills add https://github.com/aspenkit/aspens --skill claude-runner

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Streamlines the Claude/Codex prompt lifecycle by providing a dedicated execution layer that loads prompts, parses stream-json output, sanitizes paths, writes skill files, and generates skill rules.

Core Features & Use Cases

  • Robust prompt loading and execution for Claude and Codex with stream-json handling
  • Safe file output extraction, path sanitization, and on-demand skill-file I/O
  • Skill rules generation and validation to keep skills in sync with code changes

Quick Start

Configure and run the claude-runner to manage prompts and skill I/O for CLI-based execution workflows.

Frequently Asked Questions about claude-runner

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

FAQPage Schema
How do I execute Claude or Codex prompts via a CLI and handle stream-json output?▼

You can execute Claude or Codex prompts via a CLI execution layer that loads prompts, streams output using the stream-json protocol, and produces deterministic skill files with safe I/O and path handling.

What is the best way to generate skill rules and keep them in sync with code changes?▼

Generating skill rules and validating them during CLI prompt execution keeps skills in sync with code changes by extracting safe file output and applying secure prompt sanitation.

How do I sanitize file paths when writing skill files from prompt-driven workflows?▼

Sanitizing file paths when writing skill files is handled automatically by the execution layer through secure prompt sanitation and safe file output extraction during prompt-driven workflows.

Does the CLI execution layer enforce capability checks for Codex prompts?▼

Yes, the CLI execution layer enforces Codex capability checks and integration constraints like the stream-json protocol to ensure safe I/O and deterministic skill file generation.

What limitations exist when parsing stream-json output from Claude or Codex prompts?▼

Parsing stream-json output requires strict adherence to the stream-json protocol and Codex capability checks; failure to meet these integration constraints will halt deterministic skill file generation.

Can I use claude-runner to test prompt-driven skills without manual file extraction?▼

Yes, you can test prompt-driven skills without manual file extraction because the execution layer automatically parses stream-json output, sanitizes paths, and writes skill files on demand.

Related Skills

Explore More Software Engineering Skills→
korchasakorchasa
1

flowai-skill-cursor-agent-integration

Interact with the cursor-agent CLI for autonomous task execution and benchmarking.

Community
Intermediate
cjinzycjinzy
2

manage-skills

Analyze code changes and update verification skills for coverage gaps.

Community
Advanced
syhyasyhya
19

file-processing

Parse, validate, and transform CSV and JSON data files.

Community
Intermediate
lucasfthlucasfth

ideation

Generate three concrete project ideas from constraint-driven prompts.

Community
Intermediate