Can I use my own domain-specific terminology?

Yes. The skill reads from 'ops/derivation-manifest.md' to map its output to your domain-native terms, ensuring consistency with your existing project vocabulary.

What is the difference between the Pipeline and /ralph?

/ralph is the underlying engine that processes individual tasks, while the Pipeline is a convenience wrapper that orchestrates the entire lifecycle including seeding, batching, and archiving.

What happens if the pipeline is interrupted during execution?

The pipeline is fully resumable. It uses a stateful queue file to track the progress of every task, allowing you to restart and pick up exactly where the process stopped.

How do I trigger the full end-to-end process?

You can trigger the skill by using the command '/pipeline [file]' or by asking Claude to 'process this end to end' or 'run the full pipeline'.

How does the pipeline handle large source files?

The extraction phase (Phase 2) automatically handles reduction and chunking, breaking down large sources into manageable claims that are processed individually by sub-agents.

Source Processing Pipeline

Name: Source Processing Pipeline
Author: LopeWale

byLopeWale

0•

Content Management

Automates the complete lifecycle of source processing from initial seeding and claim extraction to verification and archiving.

The Pipeline skill provides a robust orchestration layer for processing source files through a rigorous multi-phase workflow. It manages the entire transition from raw input to verified knowledge units by chaining seeding, claim reduction, reflection, reweaving, and final archiving. Designed for high-integrity information processing, it ensures stateful tracking via a queue system, supports domain-specific vocabulary mapping, and allows for seamless resumption if interrupted, making it an essential tool for structured knowledge management, research analysis, and complex documentation projects.

Key Features

01Domain-native vocabulary mapping via derivation manifests.

02Automatic batch summary generation and archival organization.

03Automated end-to-end workflow orchestration from Seed to Archive.

04Stateful processing with built-in resumability and error handling.

050 GitHub stars

06Sub-agent management for isolated claim reflection and verification.

Use Cases

01Managing high-volume data extraction tasks requiring multi-step validation and 'reweaving' of context.

02Automating complex documentation workflows that require rigorous verification of every claim.

03Transforming research papers or long-form documents into a structured, interlinked knowledge base.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lopewale/ampllabs pipeline

For use in Claude.ai and ChatGPT