OpenEvidence Selection Opener

OpenEvidence Selection Opener is a lightweight Chrome extension for clinicians, students, and evidence-based medicine readers. Select text on a whitelisted page, click Ask OpenEvidence, and the extension opens the selection in OpenEvidence right next to the current tab. With a validated Groq API key, it can also rewrite selected recall-style text into editable foreground questions before sending them to OpenEvidence.

Features

• Selection toolbar anchored above highlighted text.
• UpToDate and Google buttons to search the selection beside the current tab.
• Right-click Ask OpenEvidence context menu for pages and Chrome PDF selection fallback.
• Opens OpenEvidence beside the current tab, optionally in the background.
• Local query history with a dedicated options tab, recording each text sent to OpenEvidence (plus the original selection, prompt, and transformed text for PICO/custom rewrites).
• Whitelist-based activation, including file:///* support.
• BYOK Groq integration stored in chrome.storage.local.
• Built-in PICO rewrite flow for evidence-based medicine questions.
• Custom prompt buttons with user-defined names and prompt instructions.
• Editable LLM output panel with Copy, Ask OpenEvidence, and Close.
• Light, dark, or system appearance modes.
• No runtime framework, no remote UI dependencies.

Default Whitelist

https://ankiuser.net/study
https://www.openevidence.com/*
http://uptodate.com/*
nejm.org/*
file:///*

Entries are URL prefixes. * can be used as a wildcard.

Install From Release

1. Download the latest oe-extension-vX.Y.Z.zip from Releases.
2. Unzip it.
3. Open chrome://extensions.
4. Enable Developer mode.
5. Click Load unpacked.
6. Select the unzipped extension folder.

The release also includes a .crx artifact generated by GitHub Actions.

Install From Source

git clone https://github.com/htlin222/oe-extension.git
cd oe-extension
npm test
npm run build

Then load dist/extension from chrome://extensions.

Local PDF Support

For local PDF files opened in Chrome:

1. Open chrome://extensions.
2. Open this extension's details.
3. Enable Allow access to file URLs.

Chrome's built-in PDF viewer may not expose selection events to content scripts. When the floating toolbar does not appear, use the right-click context menu: select text, right click, choose Ask OpenEvidence.

Groq BYOK

The PICO and custom prompt buttons appear only after a Groq API key is validated.

1. Open extension options.
2. Under PICO rewrite, click Edit.
3. Paste a Groq API key.
4. Click Save.

The key is stored locally with chrome.storage.local; it is not committed, bundled, or sent anywhere except Groq API requests initiated by the extension.

Custom Prompts

Custom prompts let you create extra buttons beside Ask OpenEvidence and PICO. Each custom prompt has:

• a button name, such as Dx, Summarize, or Guideline;
• a prompt that tells Groq how to transform the selected text.

The model output is displayed in an editable panel. You can copy it or ask OpenEvidence with the edited result.

OpenEvidence URL

The extension opens selections with this URL shape:

https://www.openevidence.com/ask?query=<selected text>&configName=prod&attachments=%5B%5D&_rsc=1pln2

Release Workflow

Release packaging runs on GitHub release creation/publication:

• runs the test suite;
• builds dist/oe-extension-v<version>.zip;
• packs dist/oe-extension-v<version>.crx;
• uploads both files as release artifacts.

For stable CRX IDs across releases, add a repository secret named CRX_PRIVATE_KEY_BASE64 containing the base64-encoded Chrome extension private key. Without that secret, CI still generates a .crx, but Chrome may assign a new extension ID.

Development

npm test
npm run build

The project is intentionally plain Manifest V3 JavaScript, CSS, and HTML.

Citation

If you use this project, please cite it:

@software{lin2026oeextension,
  author = {Lin, Hsieh-Ting},
  title = {oe-extension: OpenEvidence selection helper for Chrome},
  year = {2026},
  url = {https://github.com/htlin222/oe-extension},
  version = {0.1.2}
}

License

This project is licensed under the MIT License.