conf-analysis

Conf-Abstract: Academic Conference Analysis Tool

An automated framework for crawling, extracting, and analyzing academic conference papers. This tool leverages AI agents to generate comprehensive conference reports.

📂 Project Structure

• conf-analysis: Core analysis logic and Skill definitions.
  • skills/paper-report: The main skill for generating paper reports. See SKILL.md for detailed workflow.
    • scripts/: Reusable scripts for paper extraction and BibTeX generation.
    • template/: Output templates for JSON and Markdown reports.
• 2026: Directory for conference-specific data (e.g., OOPSLA26, ASPLOS26). Each folder contains:
  • url: Target conference URL.
  • papers.json: Extracted paper metadata.
  • papers.bib: Generated BibTeX entries.
  • summary.md: Final analysis report.

🛠️ Setup

1. Prerequisites (uv)

This project uses uv for Python package and environment management. Install it via:

curl -LsSf https://astral.sh/uv/install.sh | sh

2. Environment Configuration

Copy the .env.example to .env and fill in your API keys:

cp .env.example .env
# Edit .env and add your tavilyApiKey
# ...
# export .env to shell environment
export $(grep -v '^#' .env | xargs)

3. Install Dependencies & MCP Services

Use uv to sync dependencies and run the setup script:

# Sync Python dependencies
uv sync

# Add MCP servers (ensure tavilyApiKey is exported)
./setup.sh

The setup.sh script also uses uvx to run MCP servers without manual installation.

4. Running Claude with Plugins

Use the provided wrapper script to start Claude with the conf-analysis plugin enabled:

./run-claude.sh

🚀 Usage

This project follows a flexible Meta-Process described in SKILL.md.

Step 1: Initialize Conference Directory

Create a folder for the specific conference under the 2026/ (or current year) directory.

mkdir -p 2026/OOPSLA26
echo "https://target-conference-url.com" > 2026/OOPSLA26/url

Step 2: Extract Paper List

1. Retrieve raw content (HTML/Markdown) via curl, Crawl4AI, or browser "Save as".
2. Use a sub-agent to generate extract_papers_custom.py using extract_papers.py as a reference template.
3. Run the script to generate papers.json.

Step 3: Batch Bibliography Generation

python3 conf-analysis/skills/paper-report/scripts/batch_bibtex.py 2026/OOPSLA26/papers.json --output 2026/OOPSLA26/papers.bib

Step 4: Generate Summary Report

Generate the final summary.md using the provided template.

🔧 Prerequisites

• Python 3.x
• Dependencies: crawl4ai, bibtexparser, tqdm, pydantic
• (Optional) MCP Servers for advanced search (Tavily, Semantic Scholar, etc.)