ralpher

Generate Ralph loop scaffolding for autonomous codex workflows. Use when user asks to set up a ralph loop, scaffold ralph, or configure autonomous development.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "ralpher" with this command: npx skills add fordtom/ralpher/fordtom-ralpher-ralpher

Ralph Loop Setup

Generate scaffolding for running codex in an autonomous while loop.

Usage

When asked to set up a ralph loop, create a ralph/ subdirectory and the following 4 files inside it. Replace [SPEC_FILES] with the spec filenames provided by user (use path relative to project root). Configure MAX_ITERATIONS as requested.

ralph/run.sh

#!/bin/bash
set -e

SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
MAX_ITERATIONS=${1:-50}
STATUS_FILE="$SCRIPT_DIR/status.md"

for ((i = 1; i <= MAX_ITERATIONS; i++)); do
    echo "=== Iteration $i/$MAX_ITERATIONS ==="

    codex exec "$(cat "$SCRIPT_DIR/prompt.md")" --model gpt-5.2-codex --full-auto --config model_reasoning_effort="xhigh"

    if [[ -f "$SCRIPT_DIR/summary.md" ]]; then
        echo "--- Summary: $(cat "$SCRIPT_DIR/summary.md")"
    fi

    STATUS=$(head -1 "$STATUS_FILE" | grep -oE '(Done|Blocked|In Progress)')

    if [[ "$STATUS" == "Done" ]]; then
        echo "✅ Complete after $i iterations"
        exit 0
    elif [[ "$STATUS" == "Blocked" ]]; then
        echo "🚫 Blocked after $i iterations - manual intervention needed"
        exit 1
    fi

    echo "Status: $STATUS - continuing..."
done

echo "⚠️ Max iterations ($MAX_ITERATIONS) reached"
exit 1

Make executable with chmod +x ralph/run.sh.

ralph/prompt.md

Study [SPEC_FILES]
Study ralph/status.md for previous work and memory

Pick the most important incomplete task and work on it. Run the full quality gate (typecheck, lint, tests, format). If you learn something important for future work, add it to ralph/status.md. Commit when done.

Once finished, overwrite ralph/summary.md with a one-line summary.

If all tasks complete, set Status to 'Done'. If stuck on something impossible, set Status to 'Blocked'.

IMPORTANT:

- Follow any/all guidelines in AGENTS.md/CLAUDE.md
- author property-based tests or unit tests, whichever is better suited

ralph/status.md

Status: In Progress

## Memory

ralph/summary.md

Create empty.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Coding

openclaw-version-monitor

监控 OpenClaw GitHub 版本更新,获取最新版本发布说明,翻译成中文, 并推送到 Telegram 和 Feishu。用于:(1) 定时检查版本更新 (2) 推送版本更新通知 (3) 生成中文版发布说明

Archived SourceRecently Updated
Coding

ask-claude

Delegate a task to Claude Code CLI and immediately report the result back in chat. Supports persistent sessions with full context memory. Safe execution: no data exfiltration, no external calls, file operations confined to workspace. Use when the user asks to run Claude, delegate a coding task, continue a previous Claude session, or any task benefiting from Claude Code's tools (file editing, code analysis, bash, etc.).

Archived SourceRecently Updated
Coding

ai-dating

This skill enables dating and matchmaking workflows. Use it when a user asks to make friends, find a partner, run matchmaking, or provide dating preferences/profile updates. The skill should execute `dating-cli` commands to complete profile setup, task creation/update, match checking, contact reveal, and review.

Archived SourceRecently Updated