EnvDoctor
FAQ
Common questions about what EnvDoctor can diagnose, what to paste, and how diagnosis sessions are billed.
What can EnvDoctor diagnose?
EnvDoctor focuses on AI coding and web development environment failures: CLI errors, package installs, Node, Python, Git, Docker, Vercel, Postgres, auth setup, and local-to-deploy mismatches. It is not a general IT helpdesk or production incident responder.
What should I paste into a diagnosis?
Paste the failed command, the full error output, recent changes, relevant versions, and screenshot text if the error only appears visually. The more exact the context, the less EnvDoctor has to guess.
Are suggested collection commands safe?
EnvDoctor prefers read-only commands such as version checks, DNS probes, config inspection, and package manager diagnostics. Review every command before running it, and do not run commands that modify files, credentials, databases, or deployments unless you understand the impact.
When are credits spent or refunded?
One diagnosis credit is spent only after the first assistant reply succeeds and the session is saved. If the first reply cannot be produced after retries, no session is created and no credit is spent. Screenshot extraction failures also do not spend credits.
How should I handle private data?
Remove secrets, tokens, cookies, private URLs, customer data, and production environment variables before submitting. EnvDoctor needs the shape of the failure, not your credentials.
How reliable are diagnosis sessions?
EnvDoctor gives diagnostic guidance, not guaranteed fixes. It should explain its assumptions, ask for missing facts when needed, and give verification steps so you can confirm whether the repair actually worked.
ENVDOCTOR