LogRocket’s Galileo AI watches user sessions, identifies the issues that matter most, and now automatically routes them to coding agents in Cursor, Claude Code, and Codex, which trace the root cause and open a pull request automatically.
BOSTON, June 23, 2026 (GLOBE NEWSWIRE) — LogRocket, creators of the AI-first session replay and analytics platform, today announced new self-improving capabilities to LogRocket Galileo’s AI, allowing users to automatically identify customer issues and work with their coding agents to develop a fix.
Galileo AI already watches real user sessions, reviews customer feedback, and analyzes your product data. It identifies the issues that matter most, prioritized by real user impact. Now, Galileo can route those issues directly to coding agents like Cursor, Claude Code, and Codex, which trace the root cause, draft a fix, and open a pull request for your team to review.
Product and engineering teams are using AI agents to automate the manual steps that go into building and maintaining software. However, this “assembly line” of software is still limited by humans prompting their agents on what to build. Coding is no longer the bottleneck; it’s figuring out what to build.
Now, Galileo decides what user issues need fixing and sends them to your agents to fix, without relying on engineers to delegate and prioritize. LogRocket has been building towards the self-improving software era since 2016, and its new capability underscores its arrival.
How does auto-dispatching agents work?
LogRocket can now automatically route the user issues Galileo discovers to your coding agents and return a drafted fix.
Here’s how it works:
- You define the criteria that trigger an auto-dispatch: severity, surface area, and the type of issue you trust an agent to take a first pass at.
- Galileo watches everything and prioritizes the issues that matter. It watches user sessions, reads customer feedback, and tracks product changes to surface and select issues based on your criteria.
- Galileo dispatches the issue to your coding agent within Cursor, Claude Code, Codex, and more, with the full debug package attached.
- The agent does the work, reading the context, tracing the root cause, writing the fix, and opening a pull request.
Your team comes back to a drafted PR instead of a ticket in the backlog.
How are product and engineering teams using Auto-Dispatch?
Last month, LogRocket launched the LogRocket MCP, and customers including Rippling and ShipStation Global have built AI agents that resolve issues, report on feature launches, and draft responses to support tickets.
At Speedway Motors, Derek Stapleton, Software Development Team Lead, has been using LogRocket to connect observability directly to his team’s coding agents:
“LogRocket has improved our team’s ability to identify, diagnose, and resolve bugs that directly affect conversion,” Stapleton said.
“Its alerting capabilities integrate seamlessly into our engineering workflow, and its MCP integration has proven especially valuable in connecting observability with modern coding agents. LogRocket has reshaped how our team prioritizes engineering work, enabling us to focus attention on the issues with the greatest financial and customer impact.”
Kicking off the era of self-improving software
Auto-dispatching represents a key step in moving towards the era of self-improving software.
The factory floor is running on its own. Your agents aren’t waiting for someone to switch them on anymore. They stop sitting idle and start working through the problems that actually matter to your users, so your team can spend its time building great products.
There will always be a need for human, artisanal engineering. What’s being automated is the routine tasks in between: the triage, the reproduction, and small fixes for bugs you already understand.
In the past, a bug could sit in a backlog for two weeks, not because it was hard to fix but because every fix was done by hand, one issue at a time. Someone had to notice it, reproduce it, trace it, and write the fix.
Now, Galileo catches the issue from watching real user sessions, dispatches it to a coding agent with the full debug package attached, and the agent opens a pull request before anyone has triaged the ticket. The only human time spent is the part that needs human judgment: the review.
Fix the most critical issues affecting your customers
Auto-dispatching builds on Galileo, LogRocket’s AI that watches user sessions and analyzes data across the product stack to find and prioritize what’s affecting users.
Galileo pulls from:
- Session replays: Galileo watches relevant user sessions to identify friction points within the user journey.
- Customer feedback: Galileo listens to customer calls from Zoom, Gong, and other sources to capture user sentiment.
- Support tickets: Galileo reads support tickets from sources like Zendesk and Intercom to surface customer issues.
- Project management: Galileo tracks product changes in GitHub, Linear, and Jira, and analyzes A/B test results from platforms like Optimizely and Qualtrics.
Existing customers can check out the docs to learn how to dispatch coding agents from LogRocket. If you’re new to LogRocket, visit logrocket.com to get started.
About LogRocket
LogRocket is the AI-first session replay and analytics platform. Stop manually watching session replays; Galileo AI does it for you: watching every session, reading user feedback, and tracking project changes to proactively surface what’s impacting users before it affects your business. Integrate Galileo into agentic workflows via MCP across tools like Claude, ChatGPT, Slack, and more, and enable Auto-Dispatch to give Galileo the power to dispatch issues to your coding agents in Cursor, Claude Code, and Codex. LogRocket’s AI-powered platform allows its 3,000+ global customers to provide an optimal user experience every time. Founded in 2016 in Boston, MA, the company is backed by Battery Ventures, Delta-v Capital, and Matrix Partners.
