gather-session-goals

To collect work items for a collaborative coding session, you can use an interactive skill that prompts for natural language goals, automatically infers types like code or bugfix, and stores the classified list in the active session state.

The best way to kick off a collaboration session is by transforming initial natural language goals into typed work items, preparing a pending task list, and generating a consolidated design-doc Work Items section for the team.

Yes, you can infer work item types from natural language goals. The collection process automatically assigns categories such as code, bugfix, task, or unknown to each input item based on the provided descriptions.

You can write classified session goals into a design document by storing the typed work items in the active collab session state, which then prepares and generates a consolidated Work Items section for downstream skills.

Yes, you need an active collaboration session. The workflow enforces session existence before collecting initial goals, classifying them into typed work items, and writing the pending task list to the session state.

Unknown work items are explicitly assigned an unknown type during classification. The interactive collection process iterates through your items, categorizes recognized ones as code or bugfix, and prompts for confirmation before storing the final list.