What problem does it solve?
Managing cognee's AI memory from the terminal requires knowing which CLI commands map to which SDK operations, what flags each accepts, and which defaults and gotchas (dataset targeting, session caching, .env resolution) silently change behavior.
Core Features & Use Cases
- Memory Commands: Ingest text, files, or URLs with
remember, query the knowledge graph with recall, delete data with forget, and enrich graphs with improve.
- Session Memory Bridging: Read session Q&A history with
sessions get and merge session content into the graph using improve -s.
- Administration: Manage datasets, view and persist configuration to
.env, launch the UI server, and run Alembic database migrations after upgrades.
- Use Case: After upgrading cognee, run
cognee-cli upgrade to apply migrations, then cognee-cli remember ./docs --dataset-name my_project to rebuild the knowledge graph, and cognee-cli recall "question" -d my_project to query it.
Quick Start
Use the cognee-cli skill to ingest the docs folder into a dataset named my_project and then recall an answer from it.