Quordo

Security & Data Handling

The short version for your security team: there is nothing new to review, because the app sends data nowhere.

Runs on Atlassian  Questionnaire Drafter for Confluence is a Forge app that runs entirely on Atlassian's infrastructure. It has no external servers and makes no network calls outside Atlassian.

Data flow — everything stays inside Atlassian

Your browser Forge (your Atlassian instance, your region) ─────────── ──────────────────────────────────────────── questionnaire.xlsx │ parsed locally (SheetJS) │ only question TEXT leaves the page ▼ question text ─────────────▶ app resolver │ ├─▶ Confluence search + page read (your content) │ ├─▶ Atlassian-hosted LLM (Forge LLMs API) │ question + your excerpts → drafted answer │ (Atlassian-operated; no external provider) │ └─▶ Forge app storage (results + answer library) ◀───────────── drafts + citations review & export .xlsx (in browser)

No arrow leaves the Atlassian boundary. The app makes no calls to the developer or any third party.

Why this matters for the buyer

Permissions the app requests (and why)

ScopeWhy
read:page:confluenceRead the pages it quotes in answers
search:confluenceFind the pages relevant to each question
storage:appStore job results and your approved-answer library, in-instance

The app requests no write access to your Confluence content and no external egress.

The uploaded file

The questionnaire file is parsed in your browser. Its raw bytes are never sent to the app backend — only the extracted question text is processed. The file itself never reaches any server.

AI specifics

Drafting uses Atlassian-hosted Claude models via the Forge LLMs API. Inputs (the question and the relevant Confluence excerpts) are processed by that Atlassian service to produce a draft and are not used to train models. Every answer is a draft for human review by design — the app cites its sources or says "not found," rather than inventing.

Contact

Security questions: security@quordo.tech