Hello community, We’re exploring if we can govern Langdock-based AI agents with AI Control Tower. For each Langdock agent, we’d like visibility into its purpose, owner, models and data used, status, etc. Unfortunately there is no Langdock connector from ServiceNow and not much information about how to build a custom one.
Has anyone implemented AI Asset Discovery in ServiceNow AI Control Tower for Langdock? If so, what approach did you take or would recommend?
- A custom ServiceNow REST/Service Graph Connector calling Langdock’s platform APIs?
- Did this require customization on Langdock’s side, or was connecting ServiceNow directly to Langdock’s standard existing APIs sufficient? I'm particularly interested in whether those APIs provide enough metadata for automated asset discovery or whether a custom API or connector had to be developed in Langdock.
Which asset types and metadata were you able to discover, such as agents, models, prompts, skills, tools, knowledge bases, owners, status, or usage?
Were there any limitations or challenges with authentication, API schemas, pagination, or mapping assets into AI Control Tower?
