my first Argent application: a name service and it was incredibly simple to build. users can mint a transferable UTXO representing a unique name (~=NFT). they can transfer it, or co-spend it with another covenant-based application to prove ownership of that name: it's composable later. uniqueness is enforced through a compressed sparse merkle tree, preventing the same name from being minted twice while keeping computation low. the tree implementation is useful beyond this showcase, which is another reason why i wanted to share it. a worst-case sparse-tree insertion may require rebuilding up to 256 levels. @michaelsuttonil suggested exposing two registration paths: a bounded fast path for compressed proofs, and a general fallback. this reduces script size and compute budget hence network fees, in the common case. reminder that kcc-02 (and soon kcc-03) are still being discussed and so the ownership and authentication model should not be treated as final, as is, it will not be inter-operable across wallet ecosystem, which should be an imminent focus. this is a proof of concept rather than a full-featured reference name service. the goal is also to give developers and LLMs a concrete Argent example of these techniques.