JSON Storage MCP API
AI agent memory and state management via Model Context Protocol
The JSON Storage MCP API provides AI assistants with simple key-value storage for unstructured JSON data. Optimized for AI integrations using the MCP protocol over HTTP transport with JSON-RPC 2.0.
Two Endpoint Types Available:
- • General Endpoint:
/mcp/v1/storage/json- Manage multiple categories withstorage_json_*tools - • Category-Specific:
/mcp/v1/storage/json/{category}- Single category with dynamic{category}_json_*tools
MCP Endpoint
https://mcp.ainoflow.io/mcp/v1/storage/jsonTransport:
Configuration
Add MCP server in the UI dialog:
Add to your Claude Desktop configuration:
Add to your Cursor MCP configuration:
Available Tools
No parameters required
Example Response:
Useful for discovering what categories exist and navigating the data structure
category data expiresAt expiresMs Example Usage:
Returns auto-generated key like: a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6
category key data expiresAt expiresMs Example Usage:
category key patch Behavior:
- • Null values remove fields
- • Objects are recursively merged
- • Arrays and primitives replace existing values
- • Item must exist (404 if not found)
Example:
category key Example:
Returns JSON document content or error if not found
category key category page limit Returns:
Array of key objects with category, key, size, createdAt, updatedAt, expiresAt, and pagination metadata
category key Returns:
category, key, createdAt, updatedAt, expiresAt, version, etag, size, createdBy, updatedBy
Available Resources
MIME Type: application/json
MIME Type: application/json
Category-Specific Endpoint
Category-Specific Endpoint
https://mcp.ainoflow.io/mcp/v1/storage/json/{category}Benefits:
- • Simpler tool interface (no
categoryparameter) - • Dynamic tool names:
{category}_json_*instead ofstorage_json_* - • Better isolation (separate MCP servers for different purposes)
- • Reduced token usage (shorter requests and responses)
- • No tool name conflicts between multiple MCP servers
General Endpoint:
Category-Specific:
Use Cases for AI Assistants
Store AI agent settings and preferences
Cache API responses with automatic expiration
Log events without managing unique keys
Update single field without fetching full document
Persistent memory for conversation context
Discover what data is stored
Maximum document size10MBKey length200 charactersCategory length100 charactersJSON structureObject or array (no primitives)Total documentsPlan-based limit| REST API | MCP Tool |
|---|---|
POST /{category} | storage_json_create |
PUT /{category}/{key} | storage_json_upsert |
PATCH /{category}/{key} | storage_json_patch |
GET /{category}/{key} | storage_json_get |
DELETE /{category}/{key} | storage_json_delete |
GET /{category} | storage_json_list_keys |
GET /{category}/{key}/meta | storage_json_get_metadata |
Ready to integrate Storage MCP?
Sign up for free and start using JSON storage with your AI agents