Find saved working memory
Finds saved preferences, instructions, lessons, and procedures so agents adapt to how you work. Run it before starting and at key task phases.
Created by Chris Moen • Version 19 • 22 steps
Personalise your agent experience
This flow acts as a bridge between your stored preferences and your active tasks. By retrieving saved working memory, it ensures that your AI agents don't start every conversation from zero. It searches your specific memory bank for established procedures, style guides, and past lessons to make sure every output aligns with your unique way of working. You can trigger this at the start of a project or during pivotal phases to maintain consistency across your entire team.
How the memory retrieval works
The process is straightforward and relies on your existing Memory Bank settings. When you initiate the flow, it performs a targeted search across your stored data points. It looks for keywords or categories related to your current task, such as specific brand instructions or preferred document formats. Once it identifies the relevant information, it feeds those details directly into the agent’s context window. This removes the need for you to repeat instructions or manually copy and paste guardrails every time you start a new session.
Why teams use this template
Managing multiple projects often leads to fragmented results if different team members use different instructions. This template provides a single source of truth for your operational standards. It helps you scale your efforts because the AI grows smarter and more familiar with your business logic the more you use it. Instead of general responses, you receive outputs that reflect your team's specific vocabulary and technical requirements.
Key benefits of this workflow:
- Reduces manual input by pulling existing preferences automatically.
- Maintains a consistent tone of voice and adherence to brand guardrails.
- Prevents the repetition of past mistakes by recalling previously learned lessons.
- Organises complex procedures so agents follow the exact steps you’ve defined.
- Helps new team members produce high-quality work that matches established standards.
Adaptable guardrails for every task
Whether you are drafting a technical report or managing a customer support queue, this flow ensures the AI respects your boundaries. It functions as a dynamic reference tool that adapts as your business evolves. By running this flow frequently, you keep your digital workspace organised and ensure that your agents remain helpful, accurate, and aligned with your goals.
FAQ
What does the Find Saved Working Memory flow do?
This workflow searches your stored memory bank to find specific preferences, past lessons, and operational procedures. It helps your agents align with your unique style and following your established rules for any given task.
When should I run this workflow?
You should run this flow before starting a new project and at key phases during a task. This ensures your agents always have the most relevant instructions and guardrails as your work evolves.
Which services does this memory flow use?
The flow connects directly to your memory bank settings to retrieve saved data. It uses these integrated preferences to adapt agent behaviour without you needing to repeat yourself.
How can I customise how agents adapt to my work?
You can customise what the flow finds by updating your memory bank settings with new instructions or procedures. The search tags help the agent quickly identify which saved preferences apply to your current workflow.