Remedy Remedy frequently asked questions.
Product scope, controls, availability and the difference between generating a patch and verifying a repair.
What is Remedy?
Remedy is an AI Bug Fixer and broader autonomous software remediation platform designed to connect human and machine incidents with evidence, diagnosis, controlled repair, deterministic testing and production verification.
Can software report its own problems?
The expanded platform is designed to create incidents from configured SDK errors, API or health probe failures, service and container health, runtime evidence and connected telemetry.
Is Remedy Agent available today?
The Agent and broader runtime capabilities are being introduced and are labelled in development rather than generally available.
Does Remedy require unrestricted shell access?
No. The Agent design is capability-based: customers grant specific observations and named actions rather than general server control.
Does Remedy send user text directly to an AI coding agent?
No. Reporter input is sanitised and structured with authorised evidence before diagnosis and repair orchestration.
Is Remedy tied to GitHub?
No. The core model uses mapped Git repositories and exact commits. GitHub, GitLab, Forgejo, Bitbucket and generic Git are connector targets where supported.
Can Remedy repair any language?
AI reasoning can span languages, but dependable repair requires an authorised repository, reproducible environment and meaningful project-specific gates.
What makes a fix verified?
The configured deterministic post-deployment check for the original incident must succeed. A generated or merged patch alone is not a verified outcome.