research-pipeline

Automate research from idea discovery through experiment execution to narrative report generation.

Updated May 29, 2026

Search Tags:#research #workflow orchestration #arxiv #experiments #paper writing #idea discovery #auto review

Ready to run

npx skills add https://github.com/Mang30/myskills --skill research-pipeline-mang30

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: research-pipeline
Source: https://github.com/Mang30/myskills/tree/main/skills/research-pipeline
Command: npx skills add https://github.com/Mang30/myskills --skill research-pipeline-mang30

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the challenge of turning a vague research direction into an end-to-end, evidence-backed research paper draft with experiments, reviews, and a writing handoff.

Core Features & Use Cases

Full autonomous research lifecycle: runs idea discovery, experiment implementation, iterative auto-review, and produces a narrative report suitable for paper writing.
Experiment-quality guardrails: includes cross-model code review, sanity checks, crash-safe experimental execution patterns, and multi-round adversarial review loops.
Optional paper-writing handoff: generates NARRATIVE_REPORT.md (and optionally proceeds to /paper-writing) for a target venue, reducing manual coordination overhead.

Quick Start

Use the research-pipeline skill on your topic by running: /research-pipeline "your research direction".

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 paper writing?▼

Automating the research workflow from idea discovery to paper writing involves orchestrating end-to-end stages: experiment implementation, iterative auto-review, and narrative report generation. This Skill runs that full pipeline, requiring YAML-defined arguments to configure autonomous execution and generate staged outputs like NARRATIVE_REPORT.md.

What is the best way to run autonomous experiments and generate a paper draft?▼

Running autonomous experiments requires cross-model code review, crash-safe execution patterns, and multi-round adversarial review. This Skill applies those guardrails during experimentation, logs the results, and produces a narrative report suitable for a target venue submission.

How does auto review work for research experiment code?▼

Auto review for research experiment code works by applying multi-round adversarial review loops and cross-model code review. This Skill uses configurable reviewer difficulty to sanity-check and validate experiment implementation iteratively before proceeding to the narrative report generation stage.

Can I use a research pipeline for ArXiv paper submissions with conditional human checkpoints?▼

Yes, you can use a research pipeline for ArXiv paper submissions with conditional human checkpoints. This Skill supports auto-proceed logic and staged outputs, allowing you to review the IDEA_REPORT.md or experiment logs before optionally proceeding to the paper-writing handoff.

Do I need YAML configuration to start an end-to-end research workflow?▼

Yes, you need YAML configuration to start an end-to-end research workflow. This Skill requires YAML-defined arguments to set up multi-stage orchestration, configure reviewer difficulty, and manage the conditional human checkpoints needed for autonomous experimentation and report generation.

What are the limitations of autonomous research pipelines for idea discovery?▼

Limitations of autonomous research pipelines include the need for YAML-defined arguments and conditional human checkpoints to validate outputs. This Skill cannot bypass the iterative auto-review or experiment execution stages, meaning crash-safe experimental execution patterns are mandatory to prevent failure during the workflow.