kailash-mcp
Skill0 starsChanged 8 months agoApache-2.0View on GitHub
Kailash MCP — production-ready MCP server, platform server, tools, resources, transports. Use for MCP integration, platform config, security tiers.
In this repository37 other files it gives its agents.
- .claude/rules/agents.mdAGENTS.md
- .claude/guides/claude-code/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- core-sdkSkill
- dataflowSkill
- nexusSkill
- kaizenSkill
- cheatsheetsSkill
- development-guidesSkill
- nodes-referenceSkill
- workflow-patternsSkill
- deployment-gitSkill
- frontend-integrationSkill
- testing-strategiesSkill
- architecture-decisionsSkill
- code-templatesSkill
- enterprise-infrastructureSkill
- validation-patternsSkill
- gold-standardsSkill
- security-patternsSkill
- flutter-patternsSkill
- interactive-widgetsSkill
- enterprise-ai-uxSkill
- conversation-uxSkill
- uiux-design-principlesSkill
- value-auditSkill
- ai-interaction-patternsSkill
- eatp-referenceSkill
- coc-referenceSkill
- 29-pactSkill
- claude-code-patternsSkill
- error-troubleshootingSkill
- kailash-mlSkill
- kailash-alignSkill
- co-referenceSkill
- managementSkill
- projectSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 73
Tools: Call structured functions
- Distinctive: line 75
Prompts: Use pre-defined templates
- Distinctive: line 106
Use MCP when you need to:
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Kailash MCP - Model Context Protocol Integration
Production-ready MCP server implementation built into Kailash Core SDK for seamless AI agent integration.
Overview
Kailash's MCP module provides:
- Full MCP Specification: Complete implementation of Model Context Protocol
- Multiple Transports: stdio, SSE, HTTP support
- Structured Tools: Type-safe tool definitions
- Resource Management: Expose data sources to AI agents
- Authentication: Secure MCP server access
- Progress Reporting: Real-time operation status
- Testing Support: Comprehensive testing utilities
Quick Start
from kailash.mcp_server import MCPServer
# Create MCP server
server = MCPServer(name="my-server")
# Register workflow as MCP tool
@server.tool("summarize")
def summarize_tool(text: str) -> str:
"""Summarize the given text."""
# Execute workflow
workflow = create_summary_workflow()
results = runtime.execute(workflow.build())
return results["summary"]["result"]
# Run server (stdio transport by default)
server.run()Reference Documentation
Platform Server (kailash-mcp)
- mcp-platform-overview - Architecture: FastMCP, contributor pattern, transport modes
- mcp-tool-catalog - Complete list of all 26 tools by framework and security tier
- mcp-contributor-pattern - How to write a new framework contributor module
- mcp-security-tiers - T1-T4 security model, env var controls, tier escalation
- mcp-claude-code-config - Claude Code mcpServers setup (stdio + SSE), troubleshooting
- mcp-platform-map - platform_map() output schema, connection detection, debugging
- mcp-migration-guide - Migrating from MCPServer/MCPServerBase to platform server
MCP Protocol Patterns
- mcp-transports-quick - Transport configuration (stdio, SSE, HTTP)
- mcp-structured-tools - Defining MCP tools with JSON Schema
- mcp-resources - Exposing resources to agents
Security & Operations
- mcp-authentication - Authentication and authorization
- mcp-progress-reporting - Progress updates for long operations
- mcp-testing-patterns - Testing MCP servers and tools
Key Concepts
MCP Protocol
The Model Context Protocol enables AI agents to:
- Tools: Call structured functions
- Resources: Access data sources
- Prompts: Use pre-defined templates
- Sampling: Request LLM completions
Transports
MCP supports multiple transport mechanisms:
- stdio: Standard input/output (default, simplest)
- SSE: Server-Sent Events (for web clients)
- HTTP: RESTful API (for HTTP clients)
Structured Tools
Tools are type-safe functions exposed to AI agents:
- Automatic schema generation from Python type hints
- Input validation
- Error handling
- Progress reporting
Resources
Resources expose data to AI agents:
- File systems
- Databases
- APIs
- Custom data sources
When to Use This Skill
Use MCP when you need to:
- Expose workflows as tools for AI agents
- Build MCP servers for Claude Desktop or other clients
- Integrate Kailash workflows with AI assistants
- Provide structured tools to language models
- Expose resources for RAG applications
- Build custom MCP integrations
Integration Patterns
With Core SDK (Workflow Tools)
from kailash.mcp_server import MCPServer
from kailash.workflow.builder import WorkflowBuilder
server = MCPServer(name="workflow-server")
@server.tool("process_data")
def process_tool(input: str) -> dict:
workflow = WorkflowBuilder()
# Build workflow
results = runtime.execute(workflow.build())
return results["output"]["result"]With Nexus (Multi-Channel with MCP)
from nexus import Nexus
# Nexus automatically creates MCP channel
nexus = Nexus(workflows)
nexus.run() # Includes MCP serverWith DataFlow (Database Access)
from kailash.mcp_server import MCPServer
from dataflow import DataFlow
server = MCPServer(name="db-server")
db = DataFlow(...)
@server.resource("users")
def get_users():
# Expose database via MCP resource
return db.query_users()With Kaizen (Agent Tools)
from kailash.mcp_server import MCPServer
from kaizen.base import BaseAgent
server = MCPServer(name="agent-server")
@server.tool("analyze")
def analyze_tool(text: str) -> str:
agent = AnalysisAgent()
return agent(text=text).resultCritical Rules
- ✅ Use stdio transport for local development
- ✅ Define clear tool schemas with type hints
- ✅ Implement progress reporting for long operations
- ✅ Test MCP servers with real MCP clients
- ✅ Use authentication for production servers
- ❌ NEVER expose sensitive data without authentication
- ❌ NEVER skip input validation
- ❌ NEVER mock MCP protocol in tests (use real transports)
Transport Selection
| Transport | Use Case | Pros | Cons |
| stdio | Local tools, CLI | Simple, reliable | Local only |
| SSE | Web apps | Real-time updates | Complex setup |
| HTTP | APIs, services | Standard protocol | No streaming |
Version Compatibility
- Core SDK Version: 0.9.25+
- MCP Specification: Latest
- Python: 3.8+
- Transports: stdio, SSE, HTTP
Related Skills
- 01-core-sdk - Core workflow patterns
- 03-nexus - Nexus includes MCP channel
- 04-kaizen - AI agents as MCP tools
- 02-dataflow - Database resources
Support
For MCP-specific questions, invoke:
mcp-platform-specialist- Platform server, contributor plugins, security tiers, platform_mapmcp-specialist- General MCP protocol, custom servers, auth, transportstesting-specialist- MCP testing strategies
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

