OpenHive

OpenHive is a local-first, open-source system for running autonomous business agents. It describes itself as a way to hire AI agents that do the work of a GTM team, with examples focused on research, qualification, outreach, follow-up, enrichment, and data sync.

The product centers on a Queen and a colony of worker agents. Workers are given roles, skills, tools, and a scope, and the site says they can be hired, promoted, and fired as the operation changes. OpenHive also presents agent identity as versioned and editable, with bio, skills, experience, and taste, and says these identities can be audited, edited, and forked. For execution, it advertises playbooks for outbound BDR, lead generation, research, ops triage, and lead enrichment. Those playbooks include tasks such as pulling ICP leads, drafting openers, routing replies to an inbox, logging to CRM, searching Google Maps, LinkedIn, job boards, and Shopify, monitoring sources, clustering incidents, and enriching CSV lead lists.

OpenHive is built around observability and control. It shows every call, token, and dollar, allows replay of any step and diffing of decisions, and can cap costs at the colony, queen, or task level. The runtime also uses Hive cells, described as custom orchestration on KVM/Firecracker with pinned CPU and memory, sub-second cold starts, and warm-pool scheduling. The page says cheap work runs on cheap silicon and frontier work uses a GPU pool. It also notes browser-based execution for GTM tasks, with the system spawning a real browser and working pages directly.

The site offers downloads for macOS and Windows and says users can start free, with pricing shown as $0 to start. OpenHive is a tool for teams that want autonomous agents for business process work, especially GTM and operations tasks, rather than a chatbot.