The Vindaris MCP server is at https://app.vindaris.com/api/mcp and speaks Streamable HTTP. A client signs in with OAuth 2.0 and then calls tools that read and update the strategies, goals, KPIs, projects and tasks in a Vindaris workspace. This page is the technical reference for connecting a client.
https://app.vindaris.com/api/mcpcom.vindaris/vindaris in the official MCP Registry.Most clients need only the endpoint. The other values are here for clients that authenticate with a token or do not speak MCP.
https://app.vindaris.com/api/mcp over Streamable HTTP.
The server supports OAuth 2.0 with dynamic client registration. Protected-resource metadata is at https://app.vindaris.com/.well-known/oauth-protected-resource.
A personal token also works. Send it as Authorization: Bearer mcp_.... This is how Claude Code connects from the terminal.
The same actions are available over REST. The OpenAPI description is at https://app.vindaris.com/api/ai/openapi.
The server is in the official MCP Registry as com.vindaris/vindaris. See the MCP Registry entry or the Glama listing.
Two clients have their own guide: Claude, where Vindaris is a connector in the Connectors Directory, and ChatGPT, where you add Vindaris as a custom MCP server.
The sequence is the same for every client. Only the place where you enter the URL differs.
In Vindaris settings, a workspace admin turns on AI Tools (MCP) and sets each resource to off, read, write, or read and write.
Give the client the endpoint https://app.vindaris.com/api/mcp and select Streamable HTTP if it asks for a transport.
Sign in with your Vindaris account when the client starts the OAuth flow, or configure a personal token as a Bearer header.
Start with get_work_graph_alerts or list_goals to confirm the connection returns your workspace.
# Claude Code, with a personal token
claude mcp add vindaris \
--transport streamable-http https://app.vindaris.com/api/mcp \
--header "Authorization: Bearer mcp_YOUR_TOKEN"
# Project .mcp.json, OAuth on first connect
{ "mcpServers": { "vindaris": { "type": "http",
"url": "https://app.vindaris.com/api/mcp" } } }
Read tools are separate from write tools, so the permissions an admin grants decide which groups a client can use.
List and open strategies, goals, KPIs, projects and tasks, including how each aligns to the level above.list_goals get_goal list_kpis get_kpi list_projects get_project list_strategys get_strategy list_tasks get_task
Create or change a strategy, goal, KPI, project or task. Updating a KPI can record a new value.create_goal update_goal create_kpi update_kpi create_project update_project create_strategy update_strategy create_task update_task
Align work to a goal or strategy, or remove an alignment.align_entities remove_alignment
Return the prioritized items that need attention, such as at-risk goals and stalled tasks.get_work_graph_alerts
Return a structured read of the workspace for a status update or a business review.generate_summary
Post a comment on a task. The comment syncs to the connected tool, such as Jira or Asana.push_comment
Send an update reminder to the owner of an item.trigger_reminder
Turn meeting notes into tracked items aligned to the plan.ingest_content
Permissions are set per workspace. An admin turns on AI Tools (MCP) in Vindaris settings and sets each resource, meaning strategy, goal, KPI, project and task, to off, read, write, or read and write. The default is off. No tool returns workspace data until an admin grants access.
Deleting needs a separate permission and an explicit confirmation. Every call runs as the signed-in user, so a client only sees what that user can see. For limits per person, see fine-grained permissions with groups.
This page covers the server. For what agents do with it, such as reading execution risk and drafting a review, see AI agents for strategy execution.
The endpoint is https://app.vindaris.com/api/mcp. It uses the Streamable HTTP transport.
With OAuth 2.0 and dynamic client registration, so most clients need only the URL. A personal token also works, sent as an Authorization: Bearer mcp_... header.
Yes. It is listed in the official MCP Registry as com.vindaris/vindaris, and on Glama.
Any client that supports MCP over Streamable HTTP. Vindaris has client guides for Claude and for ChatGPT.
Yes. The same actions are available over REST, with an OpenAPI description at https://app.vindaris.com/api/ai/openapi.
The endpoint is https://app.vindaris.com/api/mcp. An admin turns on AI Tools (MCP) in Vindaris before a client can call a tool.
Streamable HTTP | OAuth 2.0 or a personal token