Adapting Embedding Models for Agent Capability Retrieval
This work addresses the practical problem of retrieving mixed agent capabilities (native agents, tool bundles, skills) for practitioners in open agent marketplaces, but the gains are incremental as it applies existing fine-tuning methods to a new domain.
The paper investigates whether off-the-shelf retrieval models can be adapted to match user queries to executable agent capabilities in open marketplaces, and finds that fine-tuning on a curated dataset (AgentSelect) improves retrieval on two unseen catalogs (MuleRun and ClawHub).
Open agent marketplaces list native agents, tool bundles, and reusable skill packages in the same search interface, yet practitioners still have little guidance on how to retrieve across this mixed catalog. We study whether off-the-shelf retrieval models, trained for general text retrieval, can be adapted to match user queries to executable agent capabilities, and whether the learned signal transfers beyond the benchmark used for tuning. We fine-tune three open retrieval backbones, BGE-base, KaLM-v1.5, and EasyRec, on AgentSelect, which represents marketplace-visible units as capability profiles derived from public metadata, and test transfer on two catalogs not seen during training: MuleRun native agents and a ClawHub benchmark of 50 skills with 1,000 queries. Adaptation helps on both catalogs. Code and data will be released upon publication.