Alice MCP Server

Overview

Alice MCP is a lightweight, local task‑management server built specifically for Model Context Protocol (MCP) environments. It addresses the common pain point of managing software projects while working inside AI‑powered coding assistants: developers need a reliable, isolated workspace that can be queried and manipulated by the assistant without exposing sensitive data or requiring external services. Alice solves this by running entirely on the developer’s machine, using SQLite for persistence and offering a rich set of APIs that mirror the capabilities expected by tools like Claude’s Cline.

The server delivers a full‑featured project management experience tailored for AI workflows. Projects are isolated, so each assistant session can safely create, update, or delete data without affecting other projects. Alice exposes bulk operations for tasks and architectural decision records (ADRs), allowing an assistant to create or update dozens of items in a single request. Atomicity guarantees that either all changes succeed or none do, simplifying error handling for the client. The system also tracks task status history, hierarchical epics, and contextual message logs, giving the assistant a complete view of progress and discussion threads.

Key capabilities include:

• Bulk task and decision handling with Markdown‑rich content, enabling assistants to generate comprehensive documentation in one go.
• Dynamic project planning where backlogs are AI‑queryable and adapt as tasks move through states, giving the assistant real‑time insight into priorities.
• ADR integration that ties decisions directly to the tasks that prompted them, preserving institutional knowledge and making future refactoring easier.
• Cline‑optimized startup: a single script spins up both the FastAPI backend and the MCP server, eliminating manual configuration steps for developers.

In practice, Alice empowers scenarios such as automated sprint planning, continuous architectural review, and real‑time issue triage. An AI assistant can read the current backlog, propose a new epic structure, create tasks in bulk, and record decisions—all while keeping data local and secure. For teams that rely on AI tooling for rapid iteration, Alice provides the scaffolding to keep those iterations traceable and well‑structured without external dependencies.

Because it is MCP native, Alice plugs directly into any AI workflow that understands the protocol. Developers can issue simple “create task” or “record decision” commands, and the assistant receives structured responses that can be rendered in the IDE or chat interface. The server’s emphasis on local execution and minimal setup makes it an ideal choice for privacy‑conscious projects, rapid prototyping environments, or any setting where a lightweight yet powerful project manager is needed.