What Team Foundation Server unlocks
This connector is for servers you run yourself. If your organisation is on Azure DevOps Services in the cloud, use the Azure DevOps card under the Microsoft connection instead.
Connecting an on-premises server brings in:
- projects and their recently changed work items
- work item title, description, type, state, and assignee
- git repository summaries from the same collection
How to connect Team Foundation Server
- Open Integrations in VDF AI.
- Choose Team Foundation Server.
- Enter the Server URL — for example
https://tfs.example.com/tfson TFS 2015–2018, orhttps://devops.example.comon Azure DevOps Server 2019 and later. - Enter the Project collection if it is not already part of that URL, such as
DefaultCollection. - Paste a Personal access token created under User settings → Security → Personal access tokens, scoped to Work Items (Read) and Code (Read).
VDF AI validates the token by listing your projects, then stores it encrypted against your account.
Network reachability
Unlike the cloud connectors, this one needs a network path to your server. Servers on a private network work when VDF AI is deployed inside that network — which is the normal on-premise arrangement. Loopback and link-local addresses are always refused. A server presenting a self-signed certificate requires its certificate authority to be trusted on the platform host.
Typical tasks after connecting
- summarise work item state for a project collection
- trace which repository relates to a given work item
- report on what changed across a sprint without leaving your network
Access you may approve
The token you supply carries read scopes only. Builds, pipelines, and source code are outside what this connector reads.
Reconnect or disconnect
Reconnect with a fresh token when the existing one expires or is rotated.
Disconnect to remove the stored credential entirely.
Troubleshooting
TFS rejected the personal access token
The token was truncated on copy, has expired, or lacks the two read scopes. Issue a new one.
The server did not return JSON, or an endpoint was not found
The address or collection is wrong and the server answered with a sign-in page or a 404. Open the same URL in a browser, and check whether your version needs /tfs in the path.
The host could not be resolved, or the address is not allowed
VDF AI cannot reach the server from where it runs, or the address falls in a blocked range. Your administrator will need to confirm the network path.