intent-detection

Automatic workflow routing detects user intent through keyword and context analysis to map messages to specific commands. It matches phrases describing bugs or features to a routing table, producing a single command suggestion or multiple options for the system to execute.

Intent detection for command suggestion analyzes initial user messages lacking explicit commands to identify task types like bug reports or code reviews. It applies keyword matching and context analysis to map the request to the appropriate workflow command.

To handle task descriptions without commands, apply a routing table with dynamic suggestion modes. You can lock to a single command, fork into several options for complex requests, or skip suggestion entirely for trivial edits, preventing manual clarification and misrouting.

Yes, intent analysis integrates with session start hooks and other workflow skills to ensure seamless routing at the beginning of a conversation. This captures initial messages immediately, mapping user intent to the correct command before any manual input is required.

Keyword matching limitations include dependency on clear initial messages to accurately detect intent. If a user message lacks specific context or task signals, the system may fail to map the request correctly, requiring manual clarification or resulting in workflow misrouting.

Yes, intent detection works for both bug reports and new feature requests by using a routing table that links detected intents to specific workflows. It analyzes context signals to differentiate between tasks, proposing the appropriate command for each specific scenario.