Skip to main content

Introducing Luciq Resolve Agent (Smart Resolve)

Dive deep into the Resolve Agent product and all of its features

Written by Ramy

Fixing a mobile crash usually means jumping between your crash reporter, your codebase, and your team’s chat, trying to piece together what went wrong, where it happened, and how to fix it. Resolve Agent (also known as Smart Resolve) is here to change that.

Resolve Agent is an AI-powered teammate that lives inside your Luciq dashboard. It reads your crashes, understands your source code, and generates a suggested fix, and even opens a pull request in your GitHub repository, all in a matter of minutes. Instead of spending hours on root cause analysis, your engineers get to review a ready-made fix and ship it

What the Resolve Agent does for you

Resolve Agent takes the parts of crash resolution that are repetitive and time-consuming and handles them for you:

  • Analyzes the crash. It reads the stack trace, traces it back into your code, and figures out what actually went wrong.

  • Explains the root cause. You get a clear, human-readable explanation of why the crash happened, not just where.

  • Suggests a code fix. The fix is shown as a diff, right inside the dashboard, so you can review it in seconds.

  • Opens a pull request. When you’re happy with the fix, one click creates a PR in your GitHub repository for your team to review and merge.

  • Learns from your feedback. If the first fix isn’t quite right, you can guide the agent, and it will generate an improved version.

The result: crash resolution cycles that used to take hours (or days) now take minutes, and your engineers can spend their time building new features instead of debugging old code.

Before you start

To use Resolve Agent, a few things need to be in place on your account:

  1. Enablement on your account. Resolve Agent is enabled per-account by the Luciq team. If it isn’t turned on yet, reach out to your Customer Success Manager or contact Luciq Support and ask to have it enabled. Only account admins or owners can request enablement.

  2. A connected GitHub repository. Resolve Agent needs read access to your source code to trace crashes back to the correct files and functions. This is done through the Luciq CodeLink GitHub App.

  3. A supported platform. Resolve Agent currently supports iOS (Native), Android (Native), and React Native apps, with GitHub as the source code provider.

A note on privacy: By enabling Resolve Agent, you grant Luciq permission to process crash data and application source code through its AI models. Your source code is never stored in Luciq’s backend; it is only indexed as embeddings, with no reference to the original code. Processing only happens when you actively launch the agent, and you can request to disable the feature at any time.


How to use Resolve Agent

Once the Resolve Agent is enabled and your repository is connected, you’ll see the Resolve Agent widget at the top of every crash details page in the dashboard. Here’s what a typical workflow looks like:

1. Launch the agent

Open any crash in your dashboard and click Launch Resolve Agent at the top of the page. This starts the analysis and fix generation process.

The Resolve Agent widget on the crash details page.

2. Watch it work

The agent begins by checking the crash’s stack trace for readability, then moves into fix generation. As it works, you’ll see real-time progress messages describing what it’s doing, such as reading specific files, tracing the crash's origin, and building the fix. This usually takes one to two minutes, depending on the complexity of the crash and the size of your codebase.

The Resolve Agent walking through its reasoning steps in real time.

3. Review the fix

When the agent is done, you’ll see two things in the dashboard:

  • A root cause analysis that explains, in plain language, why the crash happened.

  • A suggested code fix, shown as a diff so you can see exactly what’s being changed.

Take a moment to review the fix the same way you’d review a teammate’s pull request.

The root cause analysis and the suggested code fix displayed as a diff.

4. Provide feedback (optional)

If the fix doesn’t fully solve the problem, maybe it’s missing an edge case, or the approach doesn’t match your codebase conventions. You can give the agent feedback in your own words. It will use your input to generate an improved fix. You can iterate up to 5 times per crash.

Guiding the agent with natural-language feedback to refine the suggested fix.

5. Create a pull request

Once you’re happy with the fix, click Create pull request. Resolve Agent will open a PR directly in your connected GitHub repository. From there, your team can review the changes, run CI checks, and merge them like any other pull request.

A pull request opened by Resolve Agent in your GitHub repository.

Additional things you can do

Resolve Agent is designed to fit naturally into how your team already works. A few helpful actions to know about:

  • View previous fixes. Every fix the agent generates is saved, so your team can revisit earlier suggestions at any time.

  • Generate another fix. Not happy with the current one? Submit feedback to generate a new suggestion.

  • Start over. If a new app version is available and the old fixes no longer apply, you can delete all previous fixes and restart the analysis from scratch.

How it works (in a nutshell)


For the curious, here’s what’s happening behind the scenes:

  • Source code integration. Resolve Agent connects to your GitHub repository through the Luciq CodeLink GitHub App. It indexes your code as embeddings; no raw code is stored on Luciq’s servers.

  • Crash analysis. When you launch the agent, it examines the crash’s stack trace, cross-references it with your indexed source code, and traces the execution path to identify the root cause.

  • Fix generation. Using an agentic, multi-model architecture with Retrieval-Augmented Generation (RAG), specialized AI models handle crash data analysis, code analysis, and code generation to produce a fix that’s tailored to your codebase.

  • Feedback loop. Your feedback is fed back into the agent, which generates a refined fix that addresses your specific concerns.

  • Pull request creation. When you approve a fix, Resolve Agent uses your source code connection to open a pull request so your team can review and ship it through your normal workflow.


Compatibility

Supported

Platforms

iOS (Native), Android (Native), React Native

Source code providers

GitHub

Did this answer your question?