WORK WITH LOBSTERAI
Explain familiar preferences less often
Build and refine working preferences in LobsterAI so later tasks can draw on your writing style, output requirements and relevant context.
Start with a request like this
Keep my documents concise and lead with the conclusion.
Illustrative output. Your result depends on the task, data and model.
What to prepare
State the preferences you want to reuse, such as a report structure, writing style or work background.
What you can get
Preferences and context for future tasks that reduce repeated explanations. Continue to specify each task’s concrete requirements.
A practical way to get started
- 01
State a stable preference
Separate lasting habits from one-off instructions, such as starting weekly reports with conclusions and action items.
- 02
Check a new task
Verify that the expected format and context carry over, and correct mismatches when they appear.
- 03
Keep it current
Review memory and Agent settings when your role, project or preferences change. Remove outdated information.
Memory supports task understanding but may not be applied completely every time. Restate important constraints and recent changes in the current task.