List view

• v3.0: Exposing AutoMaintainer as an MCP Server

  Refactoring the core AutoMaintainer engines to comply with the Model Context Protocol (MCP) as a stdio/SSE server. This milestone enables external AI assistants (like Claude Desktop or Cursor) to seamlessly invoke AutoMaintainer's specialized AST routing and GitHub automation tools.

  No due date
  0% complete0 open0 closed

• v2.1: AutoMaintainer Enterprise SDK

  Packaging the LangGraph agent topology and Tree-Sitter AST mapper into an importable Python SDK (`pip install automaintainer`). Empowers enterprise teams to natively embed the autonomous agent loop into existing CI/CD pipelines, custom deployment scripts, and proprietary internal tools.

  No due date

  •0/1 issues closed
  0% complete1 open0 closed

• v2.0: The AutoMaintainer CLI

  Development of a standalone Command Line Interface (CLI). Allows developers to invoke the AutoMaintainer agent team directly from their local terminal (e.g., `am start --issue 45`). Bypasses the need for the web dashboard for advanced local workflows.

  No due date

  •0/1 issues closed
  0% complete1 open0 closed

• v1.1: Supabase & Telemetry Migration

  Migrates the ephemeral WebSocket telemetry and local JSON state files into a robust, cloud-hosted Supabase PostgreSQL database. This allows for persistent execution logs, historical dashboard analytics, and cross-session agent memory.

  No due date

  •0/1 issues closed
  0% complete1 open0 closed

• v1.0: Web IDE & AST Native Integration

  The foundational release of AutoMaintainer. Focuses on transitioning the agent architecture from simple text grepping to deep syntactic understanding using Tree-Sitter ASTs. Includes the creation of the Next.js split-pane Web IDE, real-time WebSocket agent logs, and the integration of the Architect, Visionary, Reviewer, and Implementer LangGraph state machine. (Includes work from Phases 1-4).

  No due date

  •8/13 issues closed
  61% complete5 open8 closed