Monte

Monte provides a portable persona layer for various digital tools and agents. Key features include: • Data upload for persona building • Task-specific context via CLI and API • Guides agents on thinking, decision-making, and communication • Shapes risk tolerance and output evaluation • Integrates with popular tools like Claude, Codex, and Notion Monte allows users to capture their unique taste, judgment, and goals as a durable signal. This ensures that every agent can adapt to individual preferences before starting a task, eliminating the need to repeatedly provide personal context. It focuses on offering better first responses and reducing the need for resets, making agent interactions more efficient and aligned with user expectations from the outset. The system models user personas across nine distinct dimensions, moving beyond generic settings to provide nuanced guidance. This comprehensive approach results in sharper plans, balanced tradeoffs, and improved follow-through for tasks executed by agents. Whether through a command-line interface, a programmatic API, or an embedded instruction block, Monte delivers the right contextual signal for each agent, saving time and improving output quality for developers and general users alike.