Let C be a K-comonoid.
A K-submodule𝐼≤KC is said to be a coïdeal iff #m/def/ralg/hopf
Δ(𝐼)≤K𝐼⊗C+C⊗𝐼.
and 𝜖(𝐼)=0.
Motivation
It may be unclear in what sense the above definition is dual to that of a (two-sided) ideal of a K-monoid.
The analogy is made clearer when the defining property of the latter is written