docs: add Usage section (per-task prompts + inference flags)

model card: revert description expansion

model card: capabilities, usage, limitations

model card: badge label spacing

model card: banner image + badges

assets: company banner

model card: slim company links

model card: KoreaDeep intro + DEEP Agent banner

eval results: fix mean arithmetic (76.36 = average of the five dimensions)

eval results: layout dimension per maintainer re-measurement with label-mapping fix (78.84)

eval results: layout dimension on layout_element_rule_pass_rate (official metric)

eval results: align to standalone reproduction provider (77.48 full-set single pass)

model card: condense serving and methodology sections

model card: neutral serving alias

model card: trim lineage details

KDL-Frontier-Parser-nano: weights + model card + ParseBench eval results (76.48 full-set single-pass)

initial commit