debug-headless

Capture and analyze AI CLI sessions in headless mode with Markdown transcripts.

3|4|Updated Mar 11, 2026

Search Tags:#transcript #debugging #logging #cli #ai #headless

Ready to run

npx skills add https://github.com/nicsuzor/academicOps --skill debug-headless

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debug-headless
Source: https://github.com/nicsuzor/academicOps/tree/main/archived/skills/debug-headless
Command: npx skills add https://github.com/nicsuzor/academicOps --skill debug-headless

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires uv, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables debugging of AI command-line interfaces (like Claude Code or Gemini CLI) in a non-interactive, automated environment, ensuring full capture of all output and debug logs.

Core Features & Use Cases

Headless Execution: Run AI CLI commands without manual interaction.
Full Output Capture: Records all standard output, standard error, and debug logs.
Transcript Generation: Converts raw session logs into human-readable Markdown transcripts for analysis.
Use Case: Debugging an AI agent's automated workflow that failed mid-execution by reviewing the complete, captured session transcript to pinpoint the exact error.

Quick Start

Run the debug-headless skill to execute a Claude Code command with a specific prompt and capture its output.

Frequently Asked Questions about debug-headless

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

FAQPage Schema

How do I debug an AI CLI session running in headless mode?▼

To debug an AI CLI session in headless mode, you can capture all standard output, standard error, and debug logs during non-interactive execution to review the exact failure point.

How do I convert raw AI CLI logs into readable Markdown transcripts?▼

Converting raw AI CLI logs into readable Markdown transcripts involves processing the captured session data to generate a human-readable format for detailed post-mortem analysis.

Can I run Claude Code or Gemini CLI without manual interaction to capture output?▼

Yes, you can run Claude Code or Gemini CLI without manual interaction by executing them in a non-interactive, headless mode that facilitates full output and debug log collection.

What is the best way to capture debug logs from an automated AI agent workflow?▼

The best way to capture debug logs from an automated AI agent workflow is to execute the CLI in a non-interactive environment that records all standard output and standard error comprehensively.

Do I need uv installed to execute AI CLI commands in a non-interactive environment?▼

Yes, you need uv installed as a dependency to facilitate the headless execution and comprehensive output capture of AI CLI commands for debugging purposes.