Captures demonstrated workflows as structured data — not code.
Users
3
+1 in 30 days
Rating
0.0 / 5
unchanged in 30 days
Ratings
0
0 in 30 days
Store rank
#327K
#257K among extensions
Version
3.1.0
updated 25 Jun 2026
Users
monthly, 6 months
Latest, 15 Oct 2026:3 usersMove the mouse over the chart for any month
Ratings count
monthly, 6 months
Latest, 15 Oct 2026:0 ratingsMove the mouse over the chart for any month
The Chrome Web Store rounds user counts down to the leading digit, so the user line moves in steps. The ratings count is exact and is the better signal of real adoption.
About Docent
Demonstration-driven workflow capture — data, not code.
Docent records what you do in the browser, paired with your own step-by-step context, and exports it as a structured .docent.json file. Each recording is a sequence of steps; each step pairs context with the exact actions captured — clicks, typing, navigation, uploads, and more — so the result describes what happened, in order, with full context.
Most browser recording tools produce code — Playwright scripts, Selenium tests, Puppeteer flows — and assume you want to replay it in one specific framework. Docent produces data, not code. The output makes no assumptions about what receives it or what it does with it: each dispatch payload carries a reading guide and the format's schema, so any receiving system can interpret it without prior knowledge of Docent.
TWO WAYS TO DESCRIBE A STEP
- Narration mode — describe each step in your own words; suited to an agentic/LLM consumer that interprets intent.
- Simple mode (action / validation) — tag each step as an action (do this) or a validation (check this); suited to a deterministic consumer that maps steps directly, no interpretation required.
WHAT GETS CAPTURED
- Clicks, including custom components and web elements
- Keyboard input (Enter, Escape, Tab, arrow keys)
- Text input and form changes
- File uploads
- Select / dropdown changes
- Right-clicks
- Drag and drop
- Scroll (significant scrolls only, debounced)
- Page navigation (back, forward, reload, link clicks, SPA navigation)
- Tab switches, opens, and closes
- Interactions inside iframes
Sensitive values are redacted at capture time — passwords, card / SSN / secret fields, and auth tokens in URLs are masked before anything is stored.
HOW IT WORKS
1. Open the Docent side panel in Chrome
2. Create a project and a recording — capture starts automatically
3. Describe the step — narrate it, or tag it action/validation
4. Perform the actions in the browser
5. Finish the step ("Done this step" / "Commit") — repeat for each step
6. Export a .docent.json file, or dispatch it directly to a configured endpoint
DISPATCH
Configure an HTTPS endpoint, with an optional API key, in settings. Dispatch a project from the side panel — if it holds more than one recording, send them all or pick a single recording. A confirmation step shows the endpoint, the recordings, and the step count before anything leaves your browser. The payload carries a reading guide and the format's schema, so the receiving system needs no prior knowledge of Docent.
EDITING RECORDINGS
Steps can be re-recorded, reordered, and deleted at any time before export. Full version history is preserved — nothing is permanently lost.
OPEN SOURCE
Docent is open source under the GNU General Public License v3.0.
Source: https://github.com/Arsarneq/docent