sunjongos

Paper Orchestra For Claude Code — Data skill for Claude Code

Data community

World Best Practice SCI Paper Writing Skill for Claude Code — 12-Agent Pipeline with HITL, exceeding Google PaperOrchestra & AI Scientist.

How to install Paper Orchestra For Claude Code

This entry records only its repository, not the path inside it, so there is no exact command to give. Open sunjongos/paper-orchestra-for-claude-code and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Paper Orchestra For Claude Code does

World Best Practice SCI Paper Writing Skill for Claude Code — 12-Agent Pipeline with HITL, exceeding Google PaperOrchestra & AI Scientist.

Alternatives in Data

Autonovel — An autonomous novel writing pipeline, by Hermes Agent 1.5k ★
Local News Site — Publish a local news article as a standalone Elementor page, with Google News-grade structured data and image 648 ★
Football World Cup — Football And World Cup Analytics Skills 280 ★

README

PaperOrchestra Ultra — World Best Practice SCI Paper Writing Skill for Claude Code

**The most advanced AI paper writing system ever built** — benchmarked against and exceeding Google PaperOrchestra, Sakana AI Scientist, Agent Laboratory, and Stanford ResearchAgent.

What Is This?

A Claude Code **skill** (`/paper-orchestra`) that orchestrates **12 AI agents** to write SCI-level academic papers with full Human-in-the-Loop checkpoints, multi-format output, and clinician/reviewer-quality validation.

Install it in `~/.claude/skills/paper-orchestra/SKILL.md` and invoke with `/paper-orchestra`.

Benchmark Comparison

System	Organization	HITL	Agents	Output	Novelty Analysis	Bilingual	Cost
AI Scientist	Sakana AI	✗	4-stage	PDF	Basic	✗	~$15/paper
ResearchAgent	KAIST	✗	3-tier	PDF	✗	✗	—
Agent Laboratory	Si et al.	△ Limited	3-phase	Report	✗	✗	~$2.3/paper
PaperOrchestra	Google	✗	5 agents	PDF	✗	✗	—
PaperOrchestra Ultra	Here	✓ 4 checkpoints	12 agents	HTML+DOCX+PDF	✓ Deep	✓ EN+KO	Free

**Key advantages over all existing systems:**

✅ Google PaperOrchestra: No HITL, no bilingual, no multi-format → All fixed
✅ AI Scientist: No HITL, ML-only, no clinical validation → All fixed
✅ Agent Laboratory: Requires runnable code, limited HITL → Fixed

12-Agent Pipeline

[Input: Research idea / experiment results / notes — any format]
          │
★ HITL #1: Brainstorming Confirmation
          │
╔═══════════════════════════════════════════╗
║  STAGE 2: Research Plan + Venue Intel      ║
╚═══════════════════════════════════════════╝
          │
★ HITL #2: Plan + Title + Venue Confirmation
          │
╔═══════════════════════════════════════════╗
║  STAGE 3: 12-Agent Orchestration           ║
║