Client Intelligence Hub
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Client Intelligence Hub (MCP Server)
Central source of truth for all client data. Syncs every connected platform on a schedule, normalizes it into one client workspace, and serves it to MTOS and AI agents over MCP. See ARCHITECTURE.md for the full design.
Setup
Database — apply the schema to your Supabase project: paste
db/schema.sqlinto the Supabase SQL editor and run it.Environment —
.envneeds (in addition to the existing keys):DEFAULT_TENANT_ID=<uuid of your tenant row> SCHEDULER_TICK_SECONDS=60 # optional SYNC_DISABLED=true # optional: run API without syncingSeed — create the tenant and first clients (SQL editor):
insert into tenants (name) values ('Map Ranking') returning id; -- put that id in DEFAULT_TENANT_ID, then: insert into clients (tenant_id, name, website) values ('<tenant-id>', 'Acme Plumbing', 'https://acmeplumbing.com');Connect a provider — store encrypted credentials in
connectionsand link external accounts to clients via thelink_client_accountMCP tool (or SQL). Then enable its schedule:insert into sync_schedules (tenant_id, provider, interval_minutes) values ('<tenant-id>', 'clickup', 60);Run
npm install npm run dev # tsx watch npm run build && npm start
Related MCP server: mycelium-mcp
Endpoints
Endpoint | Purpose |
| MCP (Streamable HTTP). Auth: |
| Liveness check |
Claude Code connection: claude mcp add --transport http hub http://localhost:3001/mcp --header "Authorization: Bearer <secret>"
Layout
db/schema.sql Postgres schema (canonical store)
src/core/ env, logger, db client
src/lib/crypto.ts AES-256-GCM for credentials at rest
src/connectors/ Connector SDK + implementations (clickup, gohighlevel, google/*) + stubs
src/store/ Repositories: clients, entities/metrics, assets, sync bookkeeping
src/sync/ Scheduler + sync engine + outbound push queue
src/mcp/server.ts The 16-tool MCP surface
src/http.ts Streamable HTTP transport + auth
legacy/ V1 prototype (stateless token-in-params proxy) — reference onlyAdding a connector
Create
src/connectors/<provider>.tsimplementingConnector(seeclickup.tsas the template —pullreceives decrypted credentials, identity links, and the saved cursor; returns normalized entities/metrics).Replace its stub entry in
src/connectors/registry.ts.Insert a
connectionsrow (encrypted credentials) and async_schedulesrow.
No core changes required.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Private-by-default, local-first memory/context/task orchestrator for MCP apps and agents.
Managed firm-brain MCP for law firms: Clio matters, intake, calendar, and calls over OAuth.
Agent-first social media publishing, scheduling, analytics, and unified inbox API + 16-tool MCP server across 18 social platforms (X, LinkedIn, Instagram, YouTube, TikTok, Threads, Bluesky, Reddit, Telegram, Discord, Mastodon, Facebook, Pinterest).
AI memory layer for fractional CMOs -- client-partitioned minds, meeting prep, and EOS.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceAn MCP-native enterprise operations platform that unifies team management, product catalog, order processing, knowledge base, location intelligence, analytics, and DevOps monitoring into a single server accessible from any MCP-compatible AI agent.123 npmMIT
- AlicenseNot gradedqualityCmaintenanceThe coordination layer for AI agent networks, providing persistent memory, task management, inter-agent messaging, and human oversight through native MCP tools.10 npm1MIT
- FlicenseNot gradedqualityBmaintenanceEnables AI clients to search and analyze code/docs through a central Pinecone-backed vector store, inspect project inventories and git sync status, rank work by urgency, and perform deep architectural audits via MCP.-
- AlicenseNot gradedqualityBmaintenanceExposes a self-hosted shared workspace over MCP where AI agents and their teammates read and write tasks, decisions, docs and results, so each session picks up where the last one left off. It adds hybrid cited search, per-workspace OAuth or scoped tokens, change feeds for cheap re-sync, and an append-only audit trail of every agent or human write.AGPL 3.0