Skip to content

Latest commit

 

History

History
68 lines (52 loc) · 3.35 KB

File metadata and controls

68 lines (52 loc) · 3.35 KB

SPEC.md — graphmemory

Full documentation: docs/

What it is

An MCP server that turns a project directory into a queryable semantic knowledge base. Indexes markdown documentation, TypeScript/JavaScript source code, and all project files into interconnected graph structures. Supports multi-project operation, workspaces, access control, and real-time updates.

Graphs

Graph Storage Description
DocGraph docs.json Markdown chunks with cross-file links and code block extraction
CodeGraph code.json AST symbols (functions, classes, interfaces) via tree-sitter
KnowledgeGraph knowledge.json User/LLM-created notes with typed relations
FileIndexGraph file-index.json All project files with metadata and directory hierarchy
TaskGraph tasks.json Tasks with kanban workflow, priorities, assignees
SkillGraph skills.json Reusable recipes with steps, triggers, usage tracking

See docs/graphs-overview.md for data models, node IDs, and edge types.

70 MCP tools

Group Count Condition
Context 1 always
Docs + Code blocks + Cross-graph 10 docs enabled
Code 5 code enabled
File index 3 always
Knowledge 12 always
Tasks 17 always
Epics 8 task graph enabled
Skills 14 always

See docs/api-mcp.md for schemas and docs/mcp-tools-guide.md for usage guide.

Key features

Configuration

All config via graph-memory.yaml. See docs/configuration.md and graph-memory.yaml.example.

CLI

graphmemory serve --config graph-memory.yaml    # HTTP server (primary)
graphmemory index --config graph-memory.yaml -p X # index + exit
graphmemory users add --config graph-memory.yaml  # add user

See docs/cli.md for full reference.