Step-by-step guides to replicating SharePoint lists

Each guide follows one job from signing in to a scheduled run, with the steps in the order you do them. Start with the Postgres guide if you have not run SPSync before.

• SharePoint lists in LangChain, LlamaIndex, Semantic Kernel, CrewAI, AutoGen and n8n
• Sync a SharePoint list to Postgres

Your first list in Postgres

Sync a SharePoint list to Postgres works from the desktop app or the command line. You sign in, pick a site and a list along with any lists its lookups point to, and hand SPSync a connection string kept in an environment variable. After the first run you look over the tables it created, then set a schedule so later runs only pick up changes.

A replication reviewed like code

Run SPSync from CI suits teams who want a second pair of eyes on what gets copied. spsync export-config writes a YAML file describing the site's lists, a teammate reviews it in a pull request, and a pipeline in GitHub Actions or Azure DevOps runs spsync doctor and then spsync run on a timer. Connection strings and the licence key live in the pipeline's secrets, never in the file.

Agent frameworks over a replica

SharePoint lists in AI frameworks is for developers whose framework loads SharePoint files but not list rows. It covers LangChain, LlamaIndex, Semantic Kernel, CrewAI, AutoGen and n8n, then gives each one a SQL tool that reads the replica under a read-only login.

Before you start

If you are not a tenant admin, read the admin permissions guide first. Signing in as yourself reaches the sites you can open today, while unattended runs need an admin to grant Sites.Selected on the sites you name. For one destination in depth, its page in destinations covers the setup, the type mapping and the limits, and the documentation has the full reference.