GitHub is already your issue tracker. It should be your support tool too.

Frequently asked questions

What do I need to get started?

A GitHub repository with Issues enabled. We recommend a dedicated support repository so your product code stays separate. Install the Scitor GitHub App, and you'll receive a unique email address to forward your support emails to.

What happens to my data if I stop using Scitor?

Your support conversations are GitHub Issues in your own repository — they don't live on our servers, and uninstalling the app doesn't touch them. Your full ticket history stays exactly where it is, in GitHub, with no export needed. Everything Scitor itself holds — attachments, CSAT responses, contact profiles, and usage metrics — is deleted when you uninstall the GitHub App, so download any attachments you need before you do. Full detail: scitor.io/security.

How does Scitor compare to Jira Service Management?

Jira Service Management is built for internal IT support — managing requests from employees, not customers. It's optimised for the Atlassian ecosystem (Jira, Confluence) and requires significant setup. Scitor is built for customer-facing support on GitHub. Your customers email you, tickets arrive as GitHub Issues, and your team replies without leaving GitHub. No Atlassian dependency, no per-agent pricing, and setup takes under five minutes.

How do I reply to a customer?

Write a comment on the Issue with your response, then add /send on its own line. Scitor converts your Markdown to HTML and emails it to the customer. Use /sendall to reply to all participants in the thread.

Can I send emails from my own domain?

Yes. Configure a custom sending domain through your email provider so replies come from your brand address instead of a Scitor address. This improves deliverability and looks professional to your customers.

What happens to email attachments?

Incoming attachments are uploaded to secure cloud storage and linked directly in the Issue body. When replying, drag and drop files into your GitHub comment — images will be inlined and other files attached to the outgoing email.

How does the AI triage work?

Scitor runs every incoming message through AI at the edge to generate a one-line summary, detect sentiment (positive, negative, neutral), and classify the category (bug report, feature request, billing, etc.). It then creates and applies GitHub labels automatically so you can filter and prioritize without reading every message.

How is spam handled?

You can block specific senders with the /block-sender command on any Issue. Future emails from that sender are silently dropped. Your email provider's built-in spam filtering also applies before messages reach Scitor.

Does Scitor store my emails?

No. Email content passes through Scitor to create the GitHub Issue, but the message body is not stored on our servers. Only attachments are persisted in cloud storage. Your actual support conversations live in your GitHub repository — not in a third-party database. Full breakdown of what's stored and for how long: scitor.io/security.

Can I use Discussions instead of Issues?

Yes. Add a .github/scitor.yaml file to your repository with integration: discussions. Incoming messages will create Discussions instead of Issues, which is better for Q&A-style threads where community members can also help answer.

What commands are available?

Core commands: /send (reply to sender), /sendall (reply to all), /reply (send a saved reply template), /block-sender, /unblock-sender, /create-form (generate an embeddable web form), /generate-report (monthly metrics with charts), /skip-success-comment (hide confirmation messages), and /set-from-address (custom sender).

Is there a web dashboard, or do I have to use GitHub?

Both. GitHub Issues stays the source of truth — your team can keep working from the GitHub UI and everything still works. We also ship app.scitor.io, an optional support-team-shaped view: WYSIWYG reply editor with a slash menu, AI drafting, SLA timers, CRM panel showing the customer's previous tickets, and an Insights page with response time, CSAT, and SLA pass rates. Use whichever fits the moment.

How does the AI handle our customer data?

We never train on it. AI features are paid-plan only and optional — you can disable them with ai: { enabled: false } in scitor.yaml. Each AI call runs on Cloudflare Workers AI, processed at whichever Cloudflare location is handling the request rather than a fixed or guaranteed region, sends a bounded slice of context (the ticket plus a handful of retrieved knowledge-base excerpts and the agent's instruction), and the request body is not retained. See scitor.io/ai for the full transparency page, or scitor.io/security for data handling overall.

What does 'Scitor' mean?

Scitor is a Latin word meaning 'to seek to know; to ask, inquire.' It seemed fitting for a platform that helps people ask questions and find answers.