lamm-mit/skills/metabolism-init

Metabolism Initialization Skill Overview

This skill enables knowledge metabolism initialization for research topics through a structured Day 0 baseline-building process.

Key Components

Configuration Setup: The system checks for metabolism/config.json and creates it if absent, capturing research direction, keywords, categories, and processing history.

Directory Structure: Establishes organized workspace with subdirectories for knowledge, hypotheses, experiments, conversations, and logging.

Three-Phase Workflow:

1. Broad Literature Survey — Delegates to /research-collect to gather foundational and recent works without date restrictions, generating metadata and downloads

2. Paper Analysis — Extracts core methodologies and conclusions from TeX sources or PDFs, tracking processed IDs to avoid duplication

3. Knowledge State Construction — Creates indexed summaries mapping identified topics with cross-references, timelines, and open questions

Documentation: Generates metabolism/knowledge/_index.md with research goals, topic tables, and relationship mapping, plus topic-specific markdown files and timestamped initialization logs.

Operational Principles

The skill operates autonomously post-configuration, avoiding data fabrication and requiring content verification before modifications. It spawns sessions that share the working directory, streamlining collaborative processes.