research-pipeline

Automate the research lifecycle from idea discovery to manuscript submission.

1|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/Lingrongye/federated-learning --skill research-pipeline-lingrongye
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: research-pipeline
Source: https://github.com/Lingrongye/federated-learning/tree/main/Auto-claude-code-research-in-sleep/skills/research-pipeline
Command: npx skills add https://github.com/Lingrongye/federated-learning --skill research-pipeline-lingrongye

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

End-to-end automation of the complete research lifecycle, taking a bare idea through literature review, idea discovery, implementation, experiments, auto-review, and submission-ready drafting.

Core Features & Use Cases

  • End-to-end orchestration of two major workflows (idea discovery and auto-review loop) with an implementation bridge.
  • Stage-based gating (Gate 1 human checkpoint, auto-proceed options, auto-review iterations) and configurable flags (AUTO_PROCEED, ARXIV_DOWNLOAD, HUMAN_CHECKPOINT).
  • Generates outputs like IDEA_REPORT.md, AUTO_REVIEW.md, and final submission-ready summaries; supports literature ingestion, pilot-to-production scaling, and full paper drafting.

Quick Start

Invoke the pipeline with a research direction to start idea discovery and let it autonomously progress to a submission-ready manuscript.

Frequently Asked Questions about research-pipeline

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

FAQPage Schema
How do I automate the research workflow from idea discovery to manuscript submission?

You can automate the research workflow by invoking the pipeline with a research direction to start idea discovery, which then autonomously progresses through implementation, experiments, and drafting to generate a submission-ready manuscript.

Can I configure the research pipeline to require human approval before proceeding to the next stage?

Yes, the research pipeline supports stage-based gating with configurable flags like HUMAN_CHECKPOINT and AUTO_PROCEED, allowing you to require manual approval or let the workflow advance autonomously between stages.

How does the auto-review loop work for refining a research manuscript?

The auto-review loop iteratively evaluates your research implementation and drafting, applying configurable review iterations to refine the output until it meets the criteria for a submission-ready paper.

What outputs are generated when automating experiments and literature review?

Automating experiments and literature ingestion generates structured outputs including IDEA_REPORT.md for discovered concepts, AUTO_REVIEW.md for evaluation iterations, and final submission-ready summaries for your manuscript.

Does the automated research pipeline support ingesting literature from arXiv?

Yes, the pipeline supports literature ingestion during idea discovery and review, utilizing the configurable ARXIV_DOWNLOAD flag to integrate external papers into the research workflow.

What is the best way to scale a pilot research experiment into a full manuscript?

Use the implementation bridge within the pipeline to scale pilot experiments to production, orchestrating the transition from initial implementation through auto-review loops directly into full paper drafting.