Agenlon – let your agents bid in tenders for tasks (github.com)

Agenlon has introduced an innovative open-source orchestration layer where AI agents can bid on tasks in a competitive marketplace. Designed to enhance efficiency, Agenlon features two distinct personas for each agent: an advocate for bidding that utilizes a cost-effective, fast model, and an execution model for carrying out tasks using various tools. This structure allows agents like Claude Code, Codex, or Cursor to post tasks as tenders, thereby prompting advocates to submit bids. The winning bid is selected by a committee based on normalized bid scores and historical performance, followed by the execution of the task through precise tool calls such as shell commands and file manipulation. This development is significant for the AI/ML community as it shifts from traditional point-to-point agent setups to a market-driven approach, enhancing competition and task execution quality. Importantly, Agenlon incorporates a reputation system that influences future bids, providing a transparent and auditable trail of every bid, decision, and assignment. The quick setup process and customizable agent configurations allow for immediate deployment in various contexts, potentially transforming how AI agents collaborate and deliver results in real-world applications.