scientific-analysis-architect

Plan multi-phase scientific analyses with structured chapter and notebook plans.

6|Updated Feb 3, 2026

Search Tags:#workflow #orchestration #pseudocode #reproducibility #notebooks #phase-management

Ready to run

npx skills add https://github.com/dangeles/claude --skill scientific-analysis-architect

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: scientific-analysis-architect
Source: https://github.com/dangeles/claude/tree/main/claude-config/skills/scientific-analysis-architect
Command: npx skills add https://github.com/dangeles/claude --skill scientific-analysis-architect

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables teams to design and orchestrate complex, multi-phase scientific analysis workflows, aligning planning, review, and notebook generation into a reproducible process.

Core Features & Use Cases

Structured planning: Create 3-7 chapters with goals, analyses, and dependencies.
Collaborative review: Spawn expert consultants for phase-specific input and resolve conflicts with user prompts.
Notebook generation & validation: Produce per-chapter notebooks with pseudocode, validate via nbformat, and track provenance.
Resume & recovery: Persist session state to resume interrupted workflows across machines.

Quick Start

Use the scientific-analysis-architect skill to scaffold a multi-phase research workflow and generate notebook plans for a new project.

Frequently Asked Questions about scientific-analysis-architect

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

FAQPage Schema

How do I plan and orchestrate a multi-phase scientific analysis workflow?▼

You can plan multi-phase scientific analysis workflows by generating structured chapter plans that define goals, analyses, and dependencies for up to 7 chapters. This enforces phase-based validation and tracks provenance to ensure research reproducibility.

Can I generate valid Jupyter notebooks with pseudocode for my research?▼

You can generate per-chapter notebooks containing pseudocode and validate them using nbformat-compatible standards. This ensures your notebook generation maintains structural integrity and tracks provenance for each analysis phase.

How do I track provenance and ensure reproducibility across scientific notebooks?▼

Ensure reproducibility by enforcing phase-based validation during notebook generation and maintaining structured chapter plans with explicit dependencies. This aligns planning and review into a reproducible process.

Can I resume an interrupted scientific analysis workflow across different machines?▼

You can resume interrupted scientific analysis workflows across machines by persisting session state. This recovery mechanism allows you to recover multi-phase research progress without losing structured chapter plans.

How do I collaborate with experts to review phase-specific scientific analyses?▼

Collaborate with experts by spawning expert consultants for phase-specific input during scientific analysis planning. The workflow resolves conflicts through user prompts to align multi-phase research planning and review.

What is the best way to scaffold a reproducible research workflow with notebook plans?▼

The best way to scaffold a reproducible research workflow is creating 3-7 structured chapters with defined goals and dependencies, then generating per-chapter notebook plans validated via nbformat. This orchestrates planning and review into a reproducible process.