Back to Gap Map Product / UX / Orchestration & agents

Datasette Agent

open source / Overall score: 1.7

Agent plugin for Datasette, the tool for exploring and publishing structured data in SQLite. A chat interface answers natural-language questions against a database: the agent writes and runs SQL, explores the data in the background and renders results. It calls models through the datasette-llm plugin and exposes an extensible tool surface for charts and image generation.

Early-stage alpha release that depends on pre-release versions of Datasette.

Openness

5 high confidence

5.0

license: Apache-2.0(OSI)
source: public(GitHub)
core features withheld: no

The plugin is released under Apache-2.0 with its full source public on the datasette GitHub organization, and nothing is sold beside it.

https://github.com/datasette/datasette-agent recorded 2026-08-13
Apache-2.0 license badge, full plugin source published, 110 stars and 6 forks; no paid tier or enterprise directory
https://github.com/datasette/datasette-agent/blob/main/LICENSE recorded 2026-08-13
Apache License 2.0 full text
https://raw.githubusercontent.com/datasette/datasette-agent/main/LICENSE recorded 2026-09-29
The LICENSE file is the full Apache license text, beginning "Apache License" and "Version 2.0, January 2004" with "TERMS AND CONDITIONS FOR USE, REPRODUCTION".
https://raw.githubusercontent.com/datasette/datasette-agent/main/pyproject.toml recorded 2026-09-29
The pyproject shows a plain setuptools package built from the public repo: "An LLM-powered agent assistant for Datasette" with an "Apache-2.0" license field and no extras.
https://raw.githubusercontent.com/datasette/datasette-agent/main/README.md recorded 2026-09-29
The README documents the plugin as Apache-2.0 via the badge "license-Apache%202.0-blue.svg" and lists plugin extras such as "datasette-agent-charts"; searched for enterprise, paid, license key and pricing: none.

Adoption

1 high confidence

1.0

Datasette Agent is a pre-release alpha plugin, so both GitHub stars and PyPI downloads sit at the bottom of their respective scales. The two signals agree: this is an early, low-adoption plugin rather than one held back by a weak measurement.

https://github.com/datasette/datasette-agent recorded 2026-08-13
110 stars, 6 forks; alpha-stage plugin
https://pypistats.org/api/packages/datasette-agent/recent recorded 2026-09-29
"last_month":571,"last_week":107},"package":"datasette-agent","type":"recent_downloads"}

Capability

2 medium confidence

2.0

Datasette Agent is a narrow, early structured-data agent: it can explore and query a SQLite database in natural language within Datasette, but it remains an alpha release with a limited feature scope. General coding-agent benchmarks are not the right yardstick for a domain-specific plugin this early.

https://agent.datasette.io/ recorded 2026-09-29
The project site says "Ask questions about your data; the agent will write and run SQL queries to find the answers." and offers a live demo.
https://raw.githubusercontent.com/datasette/datasette-agent/main/README.md recorded 2026-09-29
The README documents built-in tools: "The agent also has a built-in `execute_write_sql` tool that can run one or more ordered write SQL statements against a mutable database." and "launch a background agent that explores the selected database or table and writes a report".
https://raw.githubusercontent.com/datasette/datasette-agent/main/README.md recorded 2026-09-29
The README lists extension plugins: "renders charts from SQL query results using Observable Plot" and "generates images using OpenAI".

Verified 2026-09-29