A comprehensive toolkit for monitoring, diagnosing, and managing Ralph orchestration loops and session diagnostic data.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ralph-operations
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install ralph-operations using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Ralph Operations?
Ralph Operations is a specialized utility designed for developers and operators working within the Ralph ecosystem. It provides deep visibility into the lifecycle of orchestration loops, enabling precise control over agent activities. By leveraging Openclaw Skills, users can analyze detailed diagnostic data, investigate backpressure triggers, and perform high-resolution post-mortem analyses of agent sessions.
This skill is essential for maintaining the health of AI coding workflows, offering tools to debug hat selection, monitor performance metrics, and manage the underlying git worktrees that facilitate parallel task execution. It transforms raw log data into actionable insights for optimizing agent efficiency.
Ralph Operations Use Cases
- Monitor active orchestration loops and follow real-time logs for live debugging.
- Debug failed merges and resolve worktree conflicts when a loop requires manual review.
- Analyze performance metrics including iteration duration and token consumption across sessions.
- Investigate agent stalls or repetitive loops by inspecting tool call patterns in the event stream.
- Perform maintenance tasks like pruning stale processes and cleaning up orphan worktrees.
How Ralph Operations Works
- The skill tracks active loop locks and session metadata stored within the .ralph directory.
- Users interact with the CLI to manage the state of loops, transitioning them through running, queued, and merged statuses.
- For background operations, it manages isolated worktrees to prevent interference with the primary workspace.
- When diagnostics are enabled, the skill captures high-fidelity logs in JSONL format, covering orchestration, agent output, and performance.
- Structured diagnostic sessions allow for deep-dive analysis using standard tools like jq to filter for errors or specific event types.
Ralph Operations Setup
To start utilizing these Openclaw Skills, ensure your environment is configured to capture diagnostic data. Enable high-verbosity tracking with the following command:
RALPH_DIAGNOSTICS=1 ralph run -p "your prompt"
You can then manage active loops and clean up your workspace using the built-in management commands:
ralph loops # List all active orchestration loops
ralph loops prune # Remove stale processes and orphan worktrees
ralph clean --diagnostics # Clear old diagnostic session data
Ralph Operations Data Schema & Taxonomy
Ralph Operations organizes data across specialized directories to maintain isolation and history. These Openclaw Skills utilize a structured schema for analysis:
| Directory/File | Description | Key Metadata |
|---|---|---|
.ralph/agent/ |
Primary loop context | Tasks, scratchpads, and session summaries |
.worktrees/<id>/ |
Isolated loop environments | Full git worktree for concurrent agents |
orchestration.jsonl |
Flow logs | Hat selection, iteration events, and backpressure |
performance.jsonl |
Resource metrics | Timing (ms) and token counts (input/output) |
errors.jsonl |
Failure logs | Error types, messages, and stack contexts |
Ralph Operations Advanced Features
- Real-time log streaming with follow mode to watch agent decision-making live.
- Interactive worktree attachment, allowing developers to shell into a loop's specific context.
- Advanced diagnostic workflows using jq for grouping errors and calculating aggregate token usage.
- Automated merge queue management for handling multiple completed agent tasks sequentially.
- Force-stop and recovery mechanisms for clearing stuck process locks and corrupted worktrees.
SKILL.md
Loading
Related Openclaw Skills
PR Demo Creation
A professional workflow for recording, converting, and validating terminal sessions into animated GIFs and SVGs for pull requests.
playwriter
A browser automation tool that executes Playwright scripts against persistent, local Chrome sessions to preserve logins and extensions.
pdd (Prompt-Driven Development)
A structured AI agent skill that converts vague project ideas into comprehensive technical specifications and step-by-step implementation plans through iterative refinement.
Find Code Tasks
A specialized utility for listing and filtering code tasks within a repository to maintain a clear overview of project status.
Ralph Tools
A foundational toolset for AI agents to interact with the local filesystem and retrieve online technical documentation.
Release Bump for ralph-orchestrator
Automates the version bumping and release triggering process for the ralph-orchestrator workspace.
Featured*
