Privacy Policy
Last updated: 2026-05-30
Speqify is a browser extension that turns a note — typed or spoken — into a structured issue and submits it to your issue tracker. Speqify has no backend and no user accounts. We do not operate any servers, we do not collect analytics or telemetry, and none of your data is ever sent to us — there is nothing to send it to.
- No Speqify servers, no accounts, no telemetry by default.
- Your settings, tracker tokens, and AI keys are stored only in your browser.
- The only data that leaves your browser goes directly to services you configure yourself — your issue tracker, and (optionally) an AI endpoint you choose.
- The microphone is used only when you turn dictation on; audio is transcribed and then discarded.
What Speqify stores, and where
All of the following is stored locally on your device and never transmitted to us:
- Settings (selected tracker, AI configuration, capture defaults, editable ticket templates) — in the browser's extension storage (
chrome.storage.local). - Tracker tokens and AI API keys that you enter — in the browser's extension storage. They are used only to authenticate the requests you trigger.
- Drafts, including any screen recording you pause — in the browser's IndexedDB. These stay on your device until you submit or delete the draft.
Uninstalling the extension removes this local data.
What Speqify accesses
- The page you are on, when you start a capture: the URL and title, an optional screenshot or screen recording, console and network errors, and a reproduction-step timeline (clicks, inputs, scrolls, navigation). This is assembled into the issue you are creating and is shown to you before anything is sent.
- The microphone, only if you enable dictation. Audio is transcribed (on-device by default) and then discarded — Speqify does not store or upload raw audio.
The only data that leaves your browser
Speqify sends data only to the third-party services you explicitly configure, and only when you act:
- Your issue tracker — Jira, GitHub, Linear, or GitLab (including self-hosted instances at a domain you provide). When you click to create an issue, Speqify sends the ticket content (and, where supported, the screenshot/recording) to that tracker's API using the token you provided. Your use of that tracker is governed by its privacy policy.
- An AI endpoint you choose (optional). By default, transcription and drafting run entirely on your device (Whisper + a small Qwen model via WebGPU/WASM). If you instead configure a remote OpenAI-compatible endpoint (e.g. OpenAI, OpenRouter, Ollama), the text/audio needed for drafting or transcription is sent to that endpoint with your key. Captured context is redacted (query strings stripped, common token shapes scrubbed) before it is sent or embedded.
On-device AI models are downloaded from a public model host (Hugging Face) only after you opt into a model tier in Settings. Speqify is not a party to any of these transfers beyond initiating the request you asked for in your browser; it receives no copy of the data.
Permissions
- Host access (
<all_urls>) lets the capture script run on any page you choose to file an issue from, and lets Speqify call your tracker/AI APIs directly (browsers otherwise block these cross-origin calls via CORS). It is not used to monitor your browsing — no browsing history or page data is collected or transmitted to us. - Side panel / tabs / scripting / storage are used to show the Speqify UI, read the current tab for context, inject the on-demand capture script, and persist your settings locally.
Children
Speqify is a developer tool and is not directed to children under 13.
Changes
If this policy changes, the "Last updated" date above will change. Material changes will be noted in the extension's changelog.
Contact
Questions or requests: open an issue at github.com/jaroslawkaczmarski/Speqify/issues.