Shipped and exploring
Roadmap
What has shipped, what is being explored, and what stays out of scope.
The promise: everything on this page stays 100% local. No servers, no accounts, no new data collection — ever. Roadmap items carry no dates. Feedback and pull requests shape what ships. Current release: v2.5.1.
Suggested next priorities
Explore local OCR and combining screenshots first. These are hypotheses to validate with user feedback, not release commitments. For recording, prioritize removing sections from the middle of a take.
Keep Express mode simple. Put additional choices in capture options, editor tools, or history without adding steps to the default capture flow. New screenshot workflows should support Chrome and Firefox; recording capabilities remain browser-dependent.
Status legend: Exploring Planned Shipped Out of scope
Capture
-
Region loupe & element snapping Exploring
A magnifier at the crosshair while you drag, with edges that snap to element boundaries.
-
Batch capture: list of URLs → one PDF Exploring
Scroll-and-stitch several pages and export them as a single multi-page PDF. It combines two parts that already exist.
Screenshot workflows
-
Readable long-page exports Exploring
A page taller than one image now splits into several images (v2.4.0). Next to explore: an overview with selected detail crops.
-
Suggested redactions Exploring
Locally highlight possible sensitive text, such as emails and phone numbers, for review and confirmation as solid redactions. Detection can miss information; it does not guarantee privacy.
-
Combine screenshots Exploring
Arrange saved captures vertically, side by side, or in a multi-page PDF with captions. Start with saved captures before adding automated batch capture.
-
Before-and-after comparison Exploring
Choose two captures from history and compare them with a slider or highlighted differences for page changes, design reviews, and bug reports.
-
Optional source caption Exploring
Add page title, URL, capture time, and viewport size to an export. Preview and remove sensitive URL parameters before sharing. This provides context, not proof of authenticity.
-
Create a visual guide Exploring
Collect captures as steps, add short captions and numbered annotations, and export a clean document locally without an account.
Recording
-
Follow-cursor zoom mode Planned
A zoom that pans to track the cursor, instead of sitting on fixed zoom blocks.
-
Platform size presets Exploring
Export at a target aspect ratio — YouTube, Shorts, square — instead of the tab’s own.
-
Split-and-delete inside a segment Exploring
Cut a segment in two on the timeline and drop the piece you do not want, instead of trimming from an end only.
Power features
-
Local OCR — copy text out of a screenshot Exploring
Select an area and copy recognized text using local OCR. Evaluate engine and language-data size, accuracy, and startup cost; load only when needed. No screenshot data leaves the device.
-
Searchable local history Exploring
Search by page title or domain, rename captures, and pin important items to retain them in history. Explore searching extracted text once local OCR is available.
Performance & polish
-
Worker-based stitch compositing Exploring
Move full-page stitching to a Web Worker with OffscreenCanvas, so the UI stays responsive on very long pages.
-
Instant editor open Exploring
Prewarm the editor bundle while the capture runs, so the tab opens near-instantly.
Out of scope
-
Capturing desktop windows or the whole screen Out of scope
It needs the
desktopCapturepermission, which adds a “capture content of your screen” install warning and a Chrome picker on every shot. OpenScreenShot ships with emptyhost_permissions, and keeping it that way is worth more than the feature.
Shipped
v2.5.1 current release
npx -y openscreenshotstarts the MCP server — with no arguments on a piped stdin, theopenscreenshotpackage starts the MCP server, as MCP clients and catalogs expect.openscreenshot servestill works, and a terminal still gets the CLI usage.
v2.5.0
- Redaction for CLI and MCP captures — the
openscreenshotCLI and MCP server cover CSS selectors, email addresses, phone numbers, API keys, and pixel rectangles with black boxes before the screenshot. The capture fails with no image if a selector matches nothing or the page moves a target during the capture.
v2.4.1
- Fix: a full-page capture now includes scroll boxes inside the page, such as tables, code blocks, and dashboard widgets, at their full height. A dashboard panel that scrolls only a little now comes out whole too.
v2.4.0
- Very long pages split into several images — a full-page capture taller than one image now saves as up to six images instead of failing. Download names each part, and the editor opens each part in its own tab.
- Capturing a local file without file access now names the toggle to turn on: “Allow access to file URLs” on the extension’s details page.
v2.3.1
- The Chrome extension’s name now includes Screen Recorder in every language, so a store search for recording finds it too.
v2.3.0
- Controls in their own tab — Record opens a control tab beside the page, with a live camera view. The controls never appear in the video.
- Record part of a tab — draw the area to record on a still of the page. Zoom, click ripples, and the pointer work inside it.
v2.2.0
- Capture a page element — hover or use the arrow keys to select a card, chart, table, or image, then grow or shrink the selection through parent and child elements.
- MP4 recording export — H.264 + AAC MP4 is now the default, with WebM beside it.
- Save into a folder inside Downloads — a
/in the filename template saves into a subfolder, and the preview shows the cleaned path. - More shapes and arrows — rounded rectangle, oval, and triangle shapes with an optional solid fill; filled, open, double, and dot arrow tips; and arrows that bend from a middle handle.
- Your own tool order — drag tools on the rail, or move the focused tool with
Alt+↑/Alt+↓. - Stroke widths that fit the capture — presets scale with the capture width, plus a new 24px preset.
- Tabbed settings — a settings page with visual controls, and the capture delay next to the screenshot modes in the popup.
- Optional error reports — after a failed capture, the popup offers a report. Nothing is sent until you press Send.
- Fixes: full-page capture now handles dashboards that scroll an inner panel, and the macOS visible-capture shortcut moved from Command+Shift+V (paste and match style) to Option+Shift+V.
v2.1.3
- Firefox rating buttons and reminders now open Mozilla reviews. Website install buttons match your browser, with an alternative for the other browser. Updated roadmap priorities and screenshot workflows.
v2.1.2
- Capture quick actions and success feedback — copy, save PNG, or export PDF from the capture progress window, with clearer clipboard and download confirmations.
v1.5.0
- Full keyboard operation and screen reader support — a focusable canvas with keyboard layer cycling, placement, move, resize, and crop; live-region announcements for every mutation; roving tabindex on both toolbars; visible focus rings and 24×24 minimum target sizes;
forced-colorsandprefers-contrastsupport; and the editor’s UI strings moved behindchrome.i18n. - Cut tool — mark one or more horizontal bands on a full-page capture and remove them with a live preview. Cuts are non-destructive and undo like every other edit.
- Named Beautify presets — one-click looks that set padding, corner radius, shadow, and background together, over the existing Beautify sliders.
- Adjustable crop handles, with undo — eight resize handles on the crop rectangle, and applying a crop no longer clears the undo/redo stack.
- Multi-select, duplicate, and arrow-key nudge — shift-click to select more than one annotation,
Alt+Dto duplicate, and arrow keys to nudge the selection. - Capture history shelf — the last twelve captures instead of one, reopenable from the editor.
- Pin a capture in a floating window — via Document Picture-in-Picture, the only always-on-top surface Chrome offers.
- Per-layer blur strength control — a slider on the blur tool instead of one fixed strength.
- Two-click recording setup —
tabCaptureis requested inline from the Record button on its first click. - Recorder undo, and a simplified rail — undo/redo for the recording editor, Beautify collapsed behind one popover, and “Add zoom” moved onto the timeline.
- Recordings, reachable from the popup — a Recordings link beside Reopen last opens the recorder’s session list, previously reachable by no button anywhere. Each row shows the date, duration, file size, source, and which tracks (mic, tab audio, webcam) it holds.
- Redesigned editor chrome — a stable style bar, a grouped tool rail, a coral-themed range slider and canvas stage that adapt to light and dark, a consistent pressed state, and real loading and error states on export.
v1.3.0
- Auto-hiding control bar — the recording control bar hides itself about three seconds after your pointer leaves it, so it stays out of the video. Move your cursor to the bottom center of the page to bring it back. It stays visible while you hover it, and while the recording is paused.
- Synthetic cursor overlay — a “Show cursor” toggle in the editor’s settings rail, on by default, draws a smooth pointer from the recorded cursor path in both the preview and the WebM export, since
tabCapturenever records the OS cursor.
v1.2.0
- Screen recorder with auto zoom — record the current tab, with a
tabCapturepermission requested once, at your first recording. The editor plays the take on a timeline with automatic zoom at cursor clicks, manual zoom blocks, and per-segment trim, then re-renders the export through canvas, mixing in mic and tab audio, click ripples, and an optional webcam bubble. 1-second chunks land in IndexedDB as you record, so a crash or a killed background page recovers.
v1.1.0
- Quick capture — an “After capture” row in the popup sends a capture to the editor, the clipboard, or your downloads folder. It covers the popup buttons, the keyboard shortcuts, and the context menu.
- Eyedropper color picker (
I) — samples the rendered canvas, so a color inside a spotlight or on a beautify gradient picks as it looks. A swatch on the color row opens Chrome’s screen-wide picker. - Drag, drop, or paste any image — annotate a picture you already have. An import takes the same seat a capture takes, so every tool and export path keeps working.
- Crash-safe editor drafts — edits are saved locally as you work and offered back when the editor reopens. Restoring is always a click.
v0.7.0
- Beautify mode — padding, rounded corners, drop shadow, and six gradient presets, a solid color, or transparent. It previews live and travels into every export.
- Export at any scale — 25 / 50 / 100 / 200 % or an exact pixel width. It refuses sizes past Chrome’s canvas limits instead of writing an empty file.
v0.6.1
- Delayed capture — Off / 3s / 5s / 10s in the popup, with a countdown on the badge, for hover states, dropdowns, and tooltips.
- Right-click context menu capture — full page, visible area, or region from the page’s own menu. It honors the delay setting.
- Repeat last region — re-capture the previous selection rectangle from the popup footer or the context menu.
v0.6.0
- Solid and mosaic redaction — a mode on the blur tool: soft blur, coarse mosaic, or an opaque fill that survives recompression.
- Spotlight tool (
O) — rectangle, rounded rectangle, or ellipse cut-outs that merge into one dim layer. - Resize every annotation type — pen and highlighter scale on their bounding box; text and step badges scale from the corner handles.
v0.5.0
- Straight line tool (
L) — the arrow without a head. - Shift constraints — square rectangles and blur regions, and lines and arrows snapped to 45°.
- Number keys
1–8set the color. - Double-click a text layer to re-edit it.
v0.4.0 and earlier
{domain}filename token, alongside{date},{time},{title},{w}, and{h}(v0.4.0).- Numbered step badges and the highlighter tool (v0.3.0).
- Shortcut sheet in the editor, with Export bound to ⌘S (v0.3.0).
- One zoom menu, with keyboard zoom controls (v0.3.0).
- Named swatches, a custom color, and recent colors (v0.3.0).
- The export dialog can save its settings as the new defaults (v0.3.0).
- Clickable filename tokens with a live preview (v0.3.0).
- Ko-fi donation links in the README, the site, and the popup footer (v0.2.7).
Shape what ships next.
What people ask for is what gets built.
Suggest a featureWant to build one of these? Pull requests are welcome. See the contributing guide.