Devin charts weekly Linear integrations through Slack

Cognition describes how it uses Devin internally for data questions. A specific example is charting cumulative Linear integrations across organizations. The team starts the request in Slack and gives the agent access to its warehouse through the Redshift MCP connection.

The reported result contains a weekly breakdown, observations and a Metabase link that can be shared. A separate section describes DANA, a specialized analysis mode that includes SQL for inspection. The article does not identify this particular chart as a DANA run or show its underlying query. The agent maintains schema knowledge and can produce visualizations without an engineer manually preparing each query.

The important boundary is the metric definition. Cognition says it learned to specify measures and time periods and to ask explicitly for helpful visualizations. That context matters when an integration count could mean cumulative enabled organizations, new events or another measure. The article does not publish the underlying warehouse data or an independently checked numerical result for this chart.

This is an internal deployment reported by the product's developer, with an identified task and shareable output. It supports a concrete data-question workflow. It does not establish a universal accuracy rate, a particular response time or permission for the agent to change the warehouse.

Prompt
Suggested prompt (written by Agents In Use, not the author's)

Source: cognition.com · The Cognition Team