aiecos-social-crm by Aiecosvietnam is an open-source Social CRM and MCP server that connects social interactions to business data. The tool synchronizes messages from Pancake platforms into a private Supabase database and exposes those records to AI agents via an MCP server for natural-language queries. Key components include a Chrome extraction extension, an admin UI with KPI cards, and a pipeline Kanban. It targets B2B distributors, D2C brands, agencies, and developers needing AI-accessible conversation histories.

What tasks can you actually use it for?

The tool functions as a social inbox, partner tracker, and lead pipeline manager by aggregating conversation histories into a database and surfacing them in an admin UI. Practical outputs include partner classification into five stages, multi-channel inbox consolidation, CSV and printable HTML exports, and a Kanban pipeline for lead management. Use cases that map directly to delivered outputs are partner triage, activity trend monitoring, and generating printable partner reports.

How reliable are AI queries and the data sync?

The built-in MCP server lets AI clients query CRM rows using natural language, with reliability tied to the sync quality. The project advertises real-time synchronization from Pancake to Supabase, so query accuracy depends on that upstream feed and timely sync. AI-generated summaries and classifications reflect the stored messages and metadata, and therefore require human verification before high-impact decisions are made.

What inputs and setup does it require?

Deployment requires a Supabase project, a Node.js or Docker host for the sync receiver and MCP server, and a Chrome-based browser for the extraction extension. A quick-start Docker configuration is included for rapid deployment. The data pipeline expects Pancake platform access, which gives direct Zalo OA, Facebook Messenger, and Instagram records; missing platform credentials prevent full ingestion.

Is it suitable for business teams and developers building assistants?

The template orientation and open-source license make it practical for developers building custom AI assistants that need conversation context and partner metadata. Administrative features such as KPI cards and activity trends support business workflows, while self-hosting on Supabase gives teams control over data residency. Compatibility with MCP-compliant clients and a Chrome extension supports integration work rather than plug-and-play use by nontechnical teams.

Best for teams prepared to operate and validate AI-driven workflows

The tool is a pragmatic option for teams that can maintain hosting and apply operational oversight to AI outputs; automated classifications and agent recommendations should be treated as draft inputs that require human review. Organizations without internal deployment capacity should plan for developer involvement, routine validation, and governance before relying on automated partner status or lead-scoring decisions.