scopes
Scopes is a next-generation local-first task and project management tool designed for symbiotic collaboration between developers and AI assistants.
claude mcp add --transport stdio kamiazya-scopes docker run -i kamiazya/scopes \ --env SCOPES_WORKDIR="Path to local scopes data directory (optional)"
How to use
Scopes is a local-first, AI-assisted task and project management tool that treats Projects, Epics, and Tasks as a unified entity called a Scope. It provides a recursive hierarchy, human-friendly aliases, and AI collaboration through comments to keep context with you as you work. Typical workflows involve creating scopes, assigning or aliasing them for quick lookup, and navigating a tree to inspect or modify nested scopes. You can search by alias, view a scope’s children, and create child scopes under a parent to build a nested workspace that always keeps context in sync with your directory focus. The MCP integration enables AI tools to participate in scope discussions by leaving comments, enabling seamless collaboration between developers and AI assistants while keeping data local and offline-capable when needed.
Available commands include creating scopes with automatic or custom aliases, linking multiple aliases to a single scope, listing or showing scopes by alias, and rendering hierarchical views. For example, you can create a scope with an auto-generated alias, add additional aliases for quick lookup, and then drill into a parent to create child scopes. The tool supports features like tab-completion when locating scopes by alias and a visual tree view to understand relationships at a glance. These capabilities are designed to work both offline and across devices, with local data staying private by design.
Beyond basic CRUD, Scopes provides workflows for workspace management, alias strategies, and AI-assisted collaboration through comments. This enables you to keep human and AI reasoning aligned within the same scope space, reducing context switching and supporting recursive project structures across your development lifecycle.
How to install
Prerequisites:
- Java Development Kit (JDK 17 or newer)
- Git
- Optional: Docker if you plan to run via containerized image
- Clone the repository:
git clone https://github.com/kamiazya/scopes.git
cd scopes
- Build the project (Gradle):
./gradlew build
This will compile the modules and produce runnable artifacts under build/ or similar output, depending on the project configuration.
- Run the server locally (example):
./gradlew run
If a packaged jar is produced, you can start it with:
java -jar build/libs/scopes-*.jar
- Alternatively, run via Docker (if you prefer containerized execution):
docker build -t kamiazya-scopes:latest .
docker run -it --rm kamiazya-scopes:latest
Notes:
- Adjust the exact Gradle task names if the project uses a custom lifecycle (e.g., ./gradlew :application:bootRun).
- Ensure your environment has network access if dependencies must be downloaded during build.
- For development, you may want to run tests with ./gradlew konsistTest or equivalent tasks as described in the project guidelines.
Additional notes
Tips and common considerations:
- The Scopes system is designed for offline-first operation; ensure local data directories are writable and backed up if needed.
- Aliases are case-insensitive and support prefix matching, so use memorable patterns like bold-tiger-x7k or auth-system to speed up lookups.
- When using MCP with AI collaborators, keep AI comments within the same scope context to minimize drift and maintain traceability.
- If you encounter issues with alias resolution, verify that the alias cache is synchronized with the latest scope definitions; you can resync or rebuild the index if provided by the tooling.
- The architecture emphasizes clean boundaries between domain layers; if you plan to extend or customize, follow the documented domain-driven design guidelines and ADRs.
Related MCP Servers
generator
CTX: a tool that solves the context management gap when working with LLMs like ChatGPT or Claude. It helps developers organize and automatically collect information from their codebase into structured documents that can be easily shared with AI assistants.
paiml -agent-toolkit
Pragmatic AI Labs MCP Agent Toolkit - An MCP Server designed to make code with agents more deterministic
mcp-bsl-platform-context
MCP сервер для AI-ассистентов (справка по синтаксису и объектной модели 1С:Предприятие)
pfsense
pfSense MCP Server enables security administrators to manage their pfSense firewalls using natural language through AI assistants like Claude Desktop. Simply ask "Show me blocked IPs" or "Run a PCI compliance check" instead of navigating complex interfaces. Supports REST/XML-RPC/SSH connections, and includes built-in complian
k8s-ai
AI-Powered Kubernetes Management System: A platform combining natural language processing with Kubernetes management. Users can perform real-time diagnostics, resource monitoring, and smart log analysis. It simplifies Kubernetes management through conversational AI, providing a modern alternative
jqassistant-graph-rag
Source code graph RAG (graphRAG) for Java/Kotlin development based on jQAssistant