import { MCPRequest, MCPResponse } from "../types/mcp"
export async function handleResourcesRead(request: MCPRequest, env: Env): Promise<MCPResponse> {
const uri = request.params?.uri
if (uri === "memory://philosophy/guide") {
return {
jsonrpc: "2.0",
id: request.id,
result: {
contents: [
{
uri: "memory://philosophy/guide",
mimeType: "text/markdown",
text: `# MCP Memory System Guide
## How It Works
The MCP Memory system stores and retrieves coding patterns, preferences, and knowledge using vector search and intelligent categorization.
### Key Features:
1. **Auto-categorization**: Memories are automatically tagged and categorized
2. **Query expansion**: Searches are expanded to find related content
3. **Relevance scoring**: Results are scored based on recency, preferences, and context
4. **Session tracking**: The system learns from your search patterns
### Slash Commands:
- \`/remember <pattern>\` - Store a new pattern or preference
- \`/search <topic>\` - Search for relevant memories
- \`/philosophy\` - Get coding philosophy for current context
### Best Practices:
1. **Search before starting**: Always search for relevant patterns before implementing
2. **Store discoveries**: When you find a pattern that works, remember it
3. **Be specific**: Include context when storing memories
4. **Use tags**: Tag memories with #language, #pattern, #preference, etc.
### Namespaces:
- \`user:*\` - Personal preferences and patterns
- \`project:*\` - Project-specific patterns
- \`system:*\` - System-managed data (expansions, scoring)
### Examples:
\`\`\`
/remember "Always use arrow functions in TypeScript" #typescript #style
/search "error handling patterns"
/philosophy testing
\`\`\`
`
}
]
}
}
}
if (uri === "memory://philosophy/recent") {
const recentMemories = await env.DB.prepare(
"SELECT content, created_at FROM memories WHERE namespace LIKE '%philosophy%' AND deleted_at IS NULL ORDER BY created_at DESC LIMIT 10"
).all()
const recentText = recentMemories.results?.map((m: any) =>
`- ${m.content.substring(0, 100)}... (${new Date(m.created_at).toLocaleDateString()})`
).join('\n') || "No recent memories found"
return {
jsonrpc: "2.0",
id: request.id,
result: {
contents: [
{
uri: "memory://philosophy/recent",
mimeType: "text/markdown",
text: `# Recent Coding Philosophy Memories\n\n${recentText}`
}
]
}
}
}
return {
jsonrpc: "2.0",
id: request.id,
error: {
code: -32602,
message: "Resource not found"
}
}
}
export function getResourcesList() {
return [
{
uri: "memory://philosophy/guide",
name: "Coding Philosophy Guide",
description: "How to use the MCP Memory system effectively",
mimeType: "text/markdown"
},
{
uri: "memory://philosophy/recent",
name: "Recent Memories",
description: "Recently added coding patterns and preferences",
mimeType: "text/markdown"
}
]
}
