What problem does it solve? Coupling a new process-based environmental model (EcoSIM, ATS, PFLOTRAN, CLM/CTSM, TEM) to the A2MC autonomous calibration framework requires building an adapter package, a source-grounded knowledge base, a RAG index, and calibration wiring — a long multi-stage arc that is easy to get wrong. This Skill is the single top-level runbook that drives that entire adapter-kit pipeline from a filled modeler questionnaire to a calibration-ready instance. ## Core Features & Use Cases - Adapter scaffolding and intake: Reads the modeler questionnaire and scaffolds the models/<name>/ adapter package (spec, parsers, version detector, backend, datasets) via init_adapter.py, with conformance validation. - Knowledge-chain orchestration: Delegates the wiki → curated seed → RAG → validate sub-chain to the knowledge-layer skills, with gates (V1–V5 validators, seed-coverage checks, milestone registration) at each stage. - Calibration wiring: Authors the per-model site-config template, canonical parameter list, validation targets, and smoke-tests the reasoning phases so the model can actually be calibrated. - Use Case: A modeler wants to run A2MC's autonomous calibration on ATS. This Skill walks them through characterizing the codebase, building and running a sample case, writing parsers, building the RAG index, and wiring the first calibration case. ## Quick Start Ask the agent to onboard your model into A2MC by providing the filled modeler questionnaire, the model source checkout, a parameter file, and a sample output file.
