Skip to content

What is the Contox MCP Server?

The Contox MCP Server (@contox/mcp-server) is a Model Context Protocol server that gives Claude and other MCP-compatible AI assistants persistent project memory. It exposes 23 tools that let the AI load, save, search, and manage structured project knowledge across sessions.

The server binary is contox-mcp and communicates over stdio using the MCP protocol.

  • Package: @contox/mcp-server
  • Binary: contox-mcp
  • Version: 3.5.0
  • Protocol: Model Context Protocol (MCP) over stdio

All 23 Tools

The MCP server provides 23 tools organized into five categories:

Memory lifecycle

ToolDescription
contox_get_memoryLoad the complete project memory brain document at session start
contox_save_sessionSave session work into persistent memory (user-initiated only)
contox_context_packBuild a focused, token-budgeted context pack for a specific task

CRUD operations

ToolDescription
contox_list_contextsList all contexts for the current project
contox_get_contextGet a specific context by ID with full content
contox_create_contextCreate a new context manually
contox_update_contextUpdate an existing context (name, description, or content)
contox_delete_contextDelete a context by ID (irreversible)

Brain management

ToolDescription
contox_searchSemantic search across all memory via V2 embeddings
contox_populateBatch create/update contexts by schemaKey
contox_add_entryAdd an entry to a journal-like context (event-sourced)
contox_get_treeGet the hierarchical brain tree
contox_approveApprove or deprecate a context (state machine)

Knowledge graph

ToolDescription
contox_get_linksGet incoming/outgoing links for a context
contox_add_linkCreate a directed link between two contexts

Operations

ToolDescription
contox_compactCompact journal entries into a summary
contox_oncall_viewView on-call operational summary and health stats
contox_explain_schemakeyDeep-dive into any brain schemaKey
contox_scanScan a codebase and push hierarchical sub-contexts
contox_git_digestRead git commits since last save for enrichment
contox_hygieneMemory hygiene agent — analyze and clean up memory items
contox_askAsk a natural-language question about your project memory
contox_auto_resolveAutomatically resolve memory items when code fixes are committed

How the AI uses these tools

A typical session follows this pattern:

  1. Session start — The AI calls contox_get_memory to load the full project brain
  2. During work — The AI uses contox_search, contox_context_pack, and other tools as needed
  3. Session end — When the user explicitly asks, the AI calls contox_save_session to persist the session's work

The MCP server handles all communication with the Contox API. The AI never needs to make HTTP requests directly.

Next steps

  • Installation — Install and verify the MCP server
  • Configuration — Configure for Claude Desktop, .mcp.json, and environment variables