debug-ci-session

Download and parse GitHub Actions artifacts to analyze Claude CI session logs.

30|5|Updated Mar 21, 2026

Search Tags:#workflow #logs #claude #debug #github-actions #ci

Ready to run

npx skills add https://github.com/max-sixty/tend --skill debug-ci-session

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debug-ci-session
Source: https://github.com/max-sixty/tend/tree/main/plugins/install-tend/skills/debug-ci-session
Command: npx skills add https://github.com/max-sixty/tend --skill debug-ci-session

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Debug Claude CI runs by downloading and parsing session log artifacts from GitHub Actions. It helps engineers quickly understand bot decisions, verify workflow behavior, and identify where a tend run diverged from expectations.

Core Features & Use Cases

Download and locate Claude session logs artifacts for a given GitHub Actions run
Parse JSONL session logs to extract messages, tool uses, and outcomes
Trace bot decision paths across a workflow to diagnose issues and verify expected behavior
Use cases include debugging a failed run, auditing a successful run, and tracing prompt injections or misconfigurations

Quick Start

Provide a run ID or URL to fetch and inspect the Claude CI session logs.

Frequently Asked Questions about debug-ci-session

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

FAQPage Schema

How do I debug a Claude CI run from GitHub Actions session logs?▼

To debug a Claude CI run, download the GitHub Actions run artifacts containing JSONL session logs, parse them to extract messages and tool uses, and trace bot decision paths to identify where the workflow diverged from expected behavior.

How do I parse JSONL session logs to trace Claude bot decisions?▼

Parse JSONL session logs by extracting individual JSON lines to isolate messages, tool uses, and outcomes, allowing you to trace the bot's decision path across the workflow and audit how the CI run behaved.

Do I need GitHub CLI to download and analyze Claude CI artifacts?▼

Yes, diagnosing Claude CI runs requires access to GitHub CLI to download GitHub Actions run artifacts, along with the ability to parse JSON lines and extract relevant fields from the session logs.

What's the best way to audit a successful Claude CI workflow run?▼

Audit a successful Claude CI run by downloading the session log artifacts from the GitHub Actions run, parsing the JSONL data to extract tool uses and outcomes, and verifying the bot's decision paths match expected behavior.

Can I trace prompt injections or misconfigurations in Claude CI sessions?▼

Yes, you can trace prompt injections or misconfigurations by downloading and parsing the Claude CI session log artifacts to analyze bot decisions, verify workflow behavior, and identify where the run diverged from expectations.

Why does my Claude CI workflow behave differently across runs?▼

Claude CI workflows can diverge across runs due to misconfigurations or unexpected bot decisions; downloading and parsing session log artifacts from GitHub Actions allows you to trace these decision paths and pinpoint the exact divergence.