Back to Home
· product · apcore-mcp
apcore-mcp
MCP and OpenAI Tools adapter for apcore modules
apcore-mcp projects registered apcore modules — including modules scanned from a complete OpenAPI 3.0 or 3.1 document via apcore-toolkit — to MCP tool definitions and OpenAI-compatible JSON definitions. It maps schemas and behavioral annotations, supports JWT validation and elicitation-based approval, and keeps execution inside the apcore registry and policy path. Deployment security still depends on the operator's transport, identity, and authorization configuration.
· 01
Features
01Adapter Boundary: Expose registered modules without duplicating business logic
02OpenAPI Import: Scan a complete OpenAPI 3.0 or 3.1 document through apcore-toolkit and expose each operation as an MCP tool
03Dual Support: Simultaneous exposure as MCP Server and OpenAI Tools
04JWT Authentication: Built-in secure token validation and identity injection
05Tool Explorer UI: Interactive browser-based testing and documentation suite
06Streaming Bridge: Real-time notification and chunk accumulation support
07User Elicitation: Interactive input requests during module execution
08Intelligent Conversion: Automated schema transformation and $ref inlining
09Annotation Mapping: Direct mapping of behavioral hints to MCP annotations
10Sanitized Execution: Secure error handling with AI suggestions and no leaks
11Dynamic Registration: Real-time module registration reflected in AI tools
· 02
Get Started
Install
$ pip install apcore-mcpQuick Start
python.pyfrom apcore import Registry
from apcore_mcp import serve
# Auto-discover all modules
registry = Registry(extensions_dir="./extensions")
registry.discover()
# Start MCP server (stdio by default)
serve(registry)
# Or export as OpenAI tools
from apcore_mcp import to_openai_tools
tools = to_openai_tools(registry)· related
Related Products
apcore
The schema-driven module development framework that apcore-mcp bridges to AI agents.
View Product
