Help us improve
Share bugs, ideas, or general feedback.
npx claudepluginhub byigitt/ralph-wiggum-windowsHow this command is triggered — by the user, by Claude, or both
Slash command
/ralph-wiggum:ralph-loopThe summary Claude sees in its command listing — used to decide when to auto-load this command
Start a self-referential development loop that iteratively improves on a task. Execute the setup script to initialize the loop: **Arguments:** $ARGUMENTS The stop hook will now intercept all exit attempts and feed the same prompt back, allowing you to see your previous work in files and iterate on improvements. **CRITICAL**: You can ONLY declare the task complete when your completion promise statement is entirely and unambiguously accurate. Do NOT output a false completion promise just to escape the loop - this would be lying. Your context persists through: - Modified files in the wo...
/ralph-loopStarts an iterative development loop that repeatedly feeds the same prompt to Claude, enabling incremental improvements across turns. Supports --max-iterations and --completion-promise flags.
/helpExplains the Ralph Wiggum technique for iterative AI development via continuous Claude loops and lists all available Ralph Wiggum plugin commands with usage examples.
/helpExplains Ralph Loop plugin for iterative AI development via repeated Claude prompts with file self-reference, and lists commands /ralph-loop to start loops and /cancel-ralph to cancel.
/ralph-loopInitializes and runs an iterative development loop with Codex review gates. Tracks progress, logs decisions, and enforces review cycles before completion.
/ralph-loopStarts Ralph Loop session with PROMPT, iterating by feeding it back using files and git history for continuity. Supports --max-iterations N and --completion-promise TEXT.
Share bugs, ideas, or general feedback.
Start a self-referential development loop that iteratively improves on a task.
Execute the setup script to initialize the loop:
powershell -NoProfile -ExecutionPolicy Bypass -File "${CLAUDE_PLUGIN_ROOT}/scripts/setup-ralph-loop.ps1" $ARGUMENTS
Arguments: $ARGUMENTS
The stop hook will now intercept all exit attempts and feed the same prompt back, allowing you to see your previous work in files and iterate on improvements.
CRITICAL: You can ONLY declare the task complete when your completion promise statement is entirely and unambiguously accurate. Do NOT output a false completion promise just to escape the loop - this would be lying.
Your context persists through: