ralph

Name: ralph
Availability: InStock
Author: shakedfb

Manages crop production tasks with customizable workflows and automated reporting for each stage.

Updated Feb 26, 2026

Search Tags:#subagents #orchestration #task-tracking #phase-management #ai-workflow #queue-processing

Ready to run

npx skills add https://github.com/shakedfb/deterministic-agentic-sdlc-workflow --skill ralph-shakedfb

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ralph
Source: https://github.com/shakedfb/deterministic-agentic-sdlc-workflow/tree/main/.claude/skills/ralph
Command: npx skills add https://github.com/shakedfb/deterministic-agentic-sdlc-workflow --skill ralph-shakedfb

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the challenge of reliably processing a backlog of queue tasks while preserving context integrity. It achieves this by spawning isolated subagents per task, ensuring fresh context for every phase and preventing cross-task leakage. It coordinates serial and parallel processing modes, enforces per-phase boundaries, and updates the queue with phase transitions and results for downstream cross-task linkage.

Core Features & Use Cases

Per-task subagent orchestration
Serial and parallel processing modes with phase sequencing
Mandatory handoff blocks and deterministic prompts
Queue re-reading and dynamic task spawning after extract tasks
Cross-batch reflect/reweave interactions and optional cross-connect validation
Auditable learnings and final report

Quick Start

Run the /ralph 3 to process three pending tasks serially.

Frequently Asked Questions about ralph

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

FAQPage Schema

How do I prevent context contamination when processing queued tasks with subagents?▼

Subagent orchestration prevents context contamination by spawning isolated subagents for each queued task, enforcing per-phase boundaries and fresh context for every execution cycle.

What is the best way to manage multi-phase AI workflows with serial and parallel task processing?▼

Multi-phase AI workflow management coordinates serial and parallel processing modes by enforcing one-phase-per-subagent execution, applying deterministic prompts, and updating the queue with phase transitions for downstream linkage.

How do I start processing a backlog of queue tasks with isolated subagents?▼

Queue task processing begins by invoking the command with the desired batch size, which processes pending tasks serially while re-reading and dynamically filtering the queue between iterations.

Can I use cross-task linking and cross-connect validation across multiple batch processing runs?▼

Cross-task linking is supported through mandatory handoff blocks and queue updates, while optional parallel cross-connect validation enables cross-batch reflect and reweave interactions across workflow phases.

When should I not use isolated subagents for task orchestration?▼

Isolated subagent orchestration is not suitable for single-task or linear workflows without a backlog, as the queue re-reading, phase sequencing, and handoff block overhead provides no benefit without batch processing.