Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Lite SQLite provides a local SQLite database layer for OpenClaw agents with low memory overhead (2-5MB typical). Supports in-memory and file-based modes, WAL mode, connection pooling, TTL-based caching, and automatic schema migration.
SQLite requires no server or configuration, making it faster to deploy for agent-local storage than alternatives like PostgreSQL or Redis.
When to use it
- Storing agent session logs between runs
- Caching API responses with expiry times
- Persisting agent memos and key-value state
- Logging structured metadata from multi-step agent tasks
- Migrating flat JSON or CSV files into queryable storage
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: store and retrieve memos for an agent across sessions
- 1Create a file-based SQLiteDB instance with WAL mode enabled
- 2Create the agent_memos table with agent_id, key, value, and expires_at columns
- 3Index agent_id and key columns for fast lookups
- 4Insert a memo with a TTL using batch_insert or insert
- 5Query memos by agent_id and key, filtering out expired rows
Memo records persisted to agent_data.db, retrievable by agent ID and key with automatic expiry handling