Does source code leave the machine?

In the local VS Code and CLI workflows, repository files are read from the user's machine or team-controlled environment. Selected context may be sent only to configured Zero Data Retention LLM APIs for inference; those providers do not retain prompts or outputs for training or later inspection. The local workflow does not upload complete repositories to Modulo. In the GitHub App workflow, FixBugs clones the target repository into its managed runtime so it can analyze the issue, generate a fix, and prepare review artifacts. Repository data is processed and stored temporarily, only for as long as needed to complete the requested bug-fix workflow.

What data goes to FixBugs?

Modulo handles subscription management, sign-in, connected-tool access, installation checks, and service metadata needed to run the product. The customer-facing account layer is not the place where complete repositories are permanently stored.

What does Zero Data Retention mean for LLM APIs?

Zero Data Retention means the selected model provider contractually does not retain prompts or outputs for training or later inspection. It is a third-party model endpoint, separate from Modulo and separate from GitHub, GitLab, Jira, or observability providers.

Can FixBugs run locally or self-hosted?

Yes. FixBugs supports local workflows through the VS Code extension and CLI agent, and self-hosted deployment is available for teams that need infrastructure control.

How is FixBugs different from coding agents?

Coding agents are optimized for writing new code from nearby context. FixBugs is built for existing bugs and production failures: gather context, rank analysis-backed hypotheses, reproduce the issue, produce a fix, and preserve reviewable artifacts.

Which integrations are supported?

FixBugs is designed for GitHub, GitLab, Jira, logs, traces, metrics, alerts, and model providers that support Zero Data Retention. Specific availability can vary by deployment and plan.

What does validation-first mean?

Validation-first means the workflow treats a diff as incomplete until it is tied to a reproduced issue, a rationale, and test or verification output that reviewers can inspect.

Security review

Questions technical buyers ask before rollout

This FAQ summarizes the evaluation points security reviewers, platform leads, and engineering managers usually need before approving a debugging workflow.

FixBugs FAQsecurity reviewtechnical buyerdata handling

Core insight

The whitepaper is designed as a buyer artifact: clear boundaries, plain-language answers, and links back to the relevant architecture topics.

FAQ

Use these answers as the short-form review layer. The architecture and privacy topics provide the deeper explanation behind the same claims.

FAQ

Continue reading

ValidationWhy a diff without reproduction is a guess.

Ready to evaluate FixBugs?

Start with the workflow surface your team already uses, then use this architecture resource for the technical review.

Read Docs Install VS Code Extension