agent_coding_eval/analysis.md

name, lang, domain, description, tags, uses_functions, uses_types, framework, entry_point, dir_path, repo_url

name	lang	domain	description	tags	uses_functions	uses_types	framework	entry_point	dir_path	repo_url
agent_coding_eval	py	datascience	Evaluacion de agentes de coding (Qwen 2.5-Coder y otros) sobre tareas reales del fn_registry.	agents coding eval qwen llm jupyter			jupyterlab	notebooks/	analysis/agent_coding_eval	https://gitea-dgg044oo04woo4ggcsws4gk0.organic-machine.com/dataforge/agent_coding_eval

Notas

Notebooks de evaluacion de agentes de coding contra tareas del registry. Prueba modelos locales (Qwen 2.5-Coder) y compara contra baselines.