Your first memory

Once your client is connected, here are some practical examples to get you started.

Storing a memory

Ask your agent to remember anything you'd want to recall later:

Remember that the production database URL is postgres://prod.example.com/myapp
and tag it with "project" and "database".

Or be more explicit:

Use the remember tool to store key="project/db-url",
value="postgres://prod.example.com/myapp", tags=["project", "database"].

Retrieving a memory

What's the production database URL?

Your agent will search its memory and call recall("project/db-url") automatically.

Practical things to store

Here are patterns that work well:

What to store	Example key	Example tags
Project conventions	project/myapp/conventions	project, myapp
Recurring decisions	decision/auth-approach	decision, architecture
Personal preferences	preferences/code-style	preferences
Team context	team/sprint-goal	team, sprint
Reference information	ref/api-endpoints	ref, api
Work in progress	wip/feature-name	wip

Listing memories by topic

Ask your agent to list everything related to a topic:

List all memories tagged "project".

Searching by meaning

You can search for memories even if you don't remember the exact key:

Search my memories for anything about database configuration.

Deleting a memory

When something is no longer relevant:

Forget the memory with key "wip/old-feature".

Tips

• Use consistent key naming — a hierarchy like project/topic/subtopic makes memories easy to recall and browse
• Tag liberally — tags are how you list related memories; a memory can have multiple tags
• Update rather than accumulate — when something changes, ask your agent to update the existing memory rather than create a new one