Cognitive Memory tools enable the Meta Agent to access and store organizational knowledge. Search through historical data, store new information, and recall relevant context for any topic.
| Tool | Description |
|---|
search_memory | Semantic search across cognitive memory |
store_memory | Save new information to memory |
recall_context | Retrieve relevant context for a topic |
list_datasets | List available memory datasets |
Search Memory
Perform semantic searches across your organization’s knowledge base.
Example Usage:
"What did we learn from the last production outage?"
"Search for documentation about the authentication flow"
"Find all runbooks related to database recovery"
"What decisions were made about the API versioning strategy?"
Store Memory
Save new information for future reference.
Example Usage:
"Store this runbook for future reference"
"Remember that the payments team prefers Slack notifications"
"Save this troubleshooting guide for Kubernetes issues"
"Record the decision to use PostgreSQL for the new service"
Stored memories become searchable immediately and can be recalled by any team member using the Meta Agent.
Recall Context
Retrieve relevant context when working on a topic.
Example Usage:
"What context do we have about the billing system?"
"Recall previous incidents involving the API gateway"
"Get background on the migration project"
List Datasets
View available memory datasets and their contents.
Example Usage:
"What datasets are available in cognitive memory?"
"Show me the contents of the runbooks dataset"
"How many items are in each memory dataset?"
Use Cases
Incident Response
"What happened during similar incidents in the past?"
"Recall the runbook for database failover"
"What was the root cause of the last API outage?"
Onboarding
"What do new engineers need to know about our CI/CD pipeline?"
"Recall the architecture overview for the platform"
"What are our coding standards for Python?"
Decision Making
"What were the reasons for choosing Kubernetes?"
"Recall discussions about the authentication provider"
"What trade-offs did we consider for the database choice?"
