Acerca de
Subconductor acts as a persistent state machine for AI agents, designed to manage complex, multi-step workflows via the Model Context Protocol (MCP). It effectively prevents 'context drift' during long refactors by externalizing the task list into a local `.subconductor/tasks.md` manifest. This approach ensures agents maintain a transparent, resumable audit trail and can dynamically handle dependencies across multiple sessions, always remembering their exact state, completed milestones, and remaining blockers.
