Dynatrace joins Cognition's Devin Oncall as launch partner for incident response

Dynatrace is a launch partner for Cognition's Devin Oncall experience, supplying the autonomous software engineer with production data so it can investigate incidents, validate deployments and assess vulnerability exposure. The integration runs through a Dynatrace extension in the Devin Marketplace.
- Dynatrace is a launch partner for Cognition's Devin Oncall, which extends the Devin autonomous software engineer to incident response.
- The Dynatrace extension is available in the Devin Marketplace and combines an MCP server with Dynatrace skills.
- Devin can access root-cause information, logs, metrics, traces, deployment events and runtime security findings, subject to configured permissions.
- Setup requires installing the extension and authenticating with a Dynatrace platform token.
- Use cases include incident investigation, deployment validation and runtime vulnerability assessment.
Why agents need production context
According to Dynatrace, autonomous agents can now generate code, propose fixes and open pull requests, but they still need a dependable way to understand what is happening in production. Devin is Cognition's autonomous software engineer. Rather than completing code one prompt at a time, it can take on whole tasks, investigate problems, make changes and open pull requests. Devin Oncall extends that approach to incident response.
Dynatrace argues that an agent working an incident needs more than source code. It has to know what changed, what is failing, where the impact is and whether a fix actually helped. The company says most agents rebuild that context on every question by querying separate data stores and matching on tags and timestamps, which costs time, accuracy and tokens.
What the Dynatrace plugin contains
The Dynatrace extension is available through the Devin Marketplace and connects Devin to a Dynatrace environment. It has two components:
- An MCP server that provides authenticated access to the Dynatrace environment.
- A collection of Dynatrace skills that teach Devin how Dynatrace data is structured and how to work with it.
Two examples named in the post are dt-dql-essentials, which covers Dynatrace Query Language syntax and query patterns, and dt-alerting, which gives guidance on anomaly detection and alerting workflows. Dynatrace says the skills load only when relevant to a task, which helps preserve context-window capacity.
Once connected, Devin can access Dynatrace problem and root-cause information, plus logs, metrics, traces, deployment events and runtime security findings, subject to the capabilities and permissions configured for the integration. To set it up, users install the extension from the plugin marketplace and authenticate with a Dynatrace platform token. According to Dynatrace, the token is stored within the user's own installation credentials and is not written into the plugin manifest.
Three use cases
Incident investigation. Dynatrace gives the example of checkout-api returning elevated 5xx errors. A user asks Devin in plain language why it is failing, and Devin pulls the affected problem, its root cause and relevant logs, metrics and traces. The user can then narrow the scope, ask what changed, and make and validate a fix in the same session. Dynatrace says the root cause comes from Dynatrace Intelligence, described as deterministic analysis over Smartscape, the live dependency graph of the environment.
Deployment validation. Before a change ships, Devin can be asked for the current failure rate, p99 latency and resource headroom of the services involved, creating a baseline. Where deployment and telemetry data are available through the integration, Devin can compare selected pre- and post-deployment signals and summarize the differences for review.
Vulnerability assessment. Before opening a pull request, users can ask whether a workload is exposed to a specific CVE. Dynatrace says it can add runtime context to vulnerability findings, which teams can use when prioritizing remediation.
What it means for teams
Dynatrace says that as agents take on more responsibility for running software, teams need evidence to verify what those agents did. In its words, "An agent working from a guess can be allowed to recommend. An agent working from fact can be allowed to act." The company also notes that Devin draws on the same Grail data that operations and SRE teams already use, so everyone works from a consistent source of truth.
The post does not give pricing or performance figures. Its claim is that grounding Devin Oncall in root-cause findings and correlated runtime context, rather than independently retrieved signals, can help the agent move from recommending a fix to acting on one with greater confidence, while giving teams telemetry to assess the outcome.


