Can AI agents automate Productboard?

Productboard is automatable today through its REST API v2. This record is the Productboard product-management workspace, including specs authored in Spark. Requests go to https://api.productboard.com/v2 with Authorization Bearer and Accept application/json. Documented REST methods are a personal access token, OAuth 2.0 authorization code, and OAuth server-to-server JWT. API tokens are created under Settings, Integrations, Public APIs, Access Token, on the Pro plan and higher. The reference lists notes, entities, members, teams, Jira and plugin integrations, and webhook subscriptions. Productboard also hosts a beta MCP server at https://mcp.productboard.com. An admin turns it on under Settings, MCP Server, and chooses which roles may connect. Sign-in is OAuth dynamic client registration with PKCE. The agent acts as the signed-in user. Documented actions are finding and reading specs, editing spec content, posting comments, and updating status. Productboard marks the server Beta. https://developer.productboard.com/llms.txt is the developer documentation index. A published language SDK and connector catalogs were not established. Computer-use viability is unassessed. API scores 7. MCP scores 6. API is the best path because tokens and server-to-server JWT cover workspace data without the MCP beta label.

Best path today: api · Overall automatability: 7/10