Stats
Actions
Tags
From step-back
Step back from a stuck fix-it loop. Analyzes what has been tried, identifies the root cause at a higher level, and confirms a diagnosis before suggesting a new direction.
How this skill is triggered — by the user, by Claude, or both
Slash command
/step-back:step-backThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
The user is stuck in a loop and needs a fresh perspective.
The user is stuck in a loop and needs a fresh perspective.
Find and read the session transcript directly:
find ~/.claude/projects/ -name "*.jsonl" -not -path "*/subagents/*" 2>/dev/null | xargs ls -t 2>/dev/null | head -1"type":"user" — those are user messages. The message.content field has the text.<command-name> tags, <system-reminder> tags, skill loading textFind the part where the user got stuck — repeated complaints without real progress. Then analyze according to your agent instructions.
npx claudepluginhub chungchihhan/step-back --plugin step-backReviews completed coding sessions to extract actionable improvements: DX friction, documentation gaps, architecture issues, anti-patterns, bug prevention, and tooling updates.
Use when repeated fix attempts fail, the agent appears stuck in a loop, or complexity is increasing without progress
Root cause analyzer for AI coding sessions. Detects scope drift, rework patterns, and root causes to improve future prompts and repo health.