ralph

Coordinate queue-based task execution with isolated subagents per phase.

Updated Feb 3, 2026
One-click install
npx skills add https://github.com/LopeWale/amplLABS --skill ralph-lopewale
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/LopeWale/amplLABS/tree/main/.claude/skill-sources/ralph
Command: npx skills add https://github.com/LopeWale/amplLABS --skill ralph-lopewale

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinates end-to-end queue-based task execution by spawning isolated subagents for every phase to prevent context leakage.

Core Features & Use Cases

  • Per-phase subagent spawning for isolated execution across serial and parallel runs
  • Batch-aware processing with cross-linking and handoff reporting
  • Strict phase progression and queue updates with mandatory subagent invocation

Quick Start

Provide N tasks to process using /ralph N, optionally filtering by batch or type, and observe the orchestration output.

Frequently Asked Questions about ralph

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

FAQPage Schema
How do I prevent context leakage when running parallel task orchestration?

Per-phase subagent spawning isolates task execution by creating dedicated subagents for every processing phase. This queue-based orchestration mechanism ensures that serial and parallel runs maintain strict context boundaries without leakage.

How do I execute batch-aware queue processing with phase progression?

Queue-based task execution coordinates batch-aware processing by applying strict phase progression across multiple queued tasks. You provide the number of tasks to process, optionally filter by batch or type, and the orchestration handles sibling linking and handoff reporting automatically.

Can I use subagent orchestration for both serial and parallel task batches?

Yes, subagent orchestration applies to both serial and parallel processing of multiple queued tasks across batches. It supports mandatory subagent invocation per task via the Task tool, ensuring isolated execution regardless of the processing mode.

What is the best way to handle cross-batch validation during task orchestration?

Cross-batch validation is an optional feature within queue-based task execution that verifies outputs across different batches. It works alongside sibling linking and handoff reporting to maintain data integrity throughout the phase progression.

Why does queue orchestration require re-reading the queue after task extraction?

Queue re-reading after extraction is a mandatory requirement to ensure accurate phase progression and task state updates. This mechanism guarantees that the orchestration processes the most current queue state before spawning the next isolated subagent.

Do I need specific dependencies to run per-phase subagent spawning?

No specific dependencies are required to run per-phase subagent spawning. The orchestration operates independently using the Task tool for mandatory subagent invocation, requiring only the input of tasks to process via the command interface.