Orcha MCP Server

What is Orcha?

Orcha is a unified context layer for AI tools and agents. It stores, indexes, and exposes organizational, team, or individual knowledge. Files, structured databases, and connected sources live in one workspace, and agents reach them through MCP, CLI with virtual file system, or REST API - with permissions, citations, and provenance.

Category: Memory & Knowledge
Author: westonhancock
Added: August 12, 2026

Why It Matters for Operators

The "context problem" is the #1 blocker for AI agents in business operations. Every session starts with amnesia - operators re-explain their business, products, customers, and processes. Orcha solves this by providing a persistent, permissioned knowledge layer that agents query directly.

This is fundamentally different from: - PLUR/AgenticMemory: Session-level memory for coding agents - Notion MCP: Document-level access to Notion pages - Groundwork: Company memory with public proof

Orcha provides a structured knowledge layer with databases, typed queries, and connected sources - not just documents, but queryable business data.

Connection Details

{
  "mcpServers": {
    "orcha": {
      "type": "streamable-http",
      "url": "https://app.tryorcha.com/mcp"
    }
  }
}

Transport: Streamable HTTP (remote)
Auth: Required (workspace-scoped)
Pricing: Not yet published (early stage)

Key Features

Feature	Description
File System	Read, create, update, organize, and delete files (permission-scoped)
Context Bundles	Curated sets of context with usage guidance (list_bundles, get_bundle)
Structured Databases	Typed queries against business data (list_databases, query_database, record CRUD)
Connected Sources	Direct read-only access to external sources (query_source, fetch_source_document)
Virtual Filesystem	Browse workspace as a tree: ls, tree, find, cat, grep, stat
Permissions	Role-based access control scoped to workspace members
Citations	Every response includes source provenance

Verified Use Cases

1. Operator Knowledge Base - Store SOPs, pricing, product specs, and customer data that agents query during every session
2. Team Context Sharing - Multiple agents across different tools (Claude, Cursor, ChatGPT) access the same organizational truth
3. Agent-Ready Business Data - Structured databases of customers, products, or deals that agents query via typed tools
4. Source-Grounded Answers - Agents cite specific files or database records in responses, eliminating hallucination

CorpusIQ Integration Opportunity

Priority: MEDIUM-HIGH. Orcha could serve as the canonical knowledge layer for CorpusIQ operators: - Product knowledge (features, pricing, roadmap) - Customer data (segments, feedback, churn signals) - Competitive intelligence (stored research, battlecards) - Content guidelines (voice, positioning, compliance)

However, we already have Honcho + GBrain for agent memory and Git-based docs for product knowledge. Orcha would be most valuable for operators who want a GUI-managed knowledge base their agents query - a different use case than our current pipeline.

Verdict

★★★★☆ Strong concept, early execution. The unified context layer approach - files + databases + sources, all permission-scoped - is the right architecture for organizational AI. Currently early-stage with thin public documentation. Worth monitoring, especially as it adds more structured database capabilities and source connectors.

Resources

• Homepage: https://www.tryorcha.com/
• Documentation: https://www.tryorcha.com/docs
• Repository: https://github.com/westonhancock/orcha-mcp
• mcp.so: https://mcp.so/servers/orcha