Python sqlite3
sqlite3 is Python's built-in DB-API 2.0 interface for SQLite. It gives agents a local, serverless database for persistence, indexing, and durable intermediate state.
Core Concepts
- SQLite stores data in a single file and does not require a separate database server.
- The Python module exposes connections, cursors, transactions, parameterized queries, and row factories.
sqlite3.Rowis useful when results should behave like mappings as well as tuples.
Significance for Agents
- Embedded storage is a strong fit for local agent memory, caches, work queues, and audit trails.
- SQLite is durable enough for many single-user or single-node workflows without adding operational overhead.
- Parameterized queries matter because prompt-derived strings should never be interpolated directly into SQL.
Practical Heuristics
- Use placeholders instead of string-built SQL.
- Decide transaction boundaries explicitly for write-heavy workflows.
- Set a row factory when downstream code benefits from named-column access.
- Treat SQLite as a local systems primitive, not a substitute for a distributed database.