SMF·GUIDDE
Can a prompt replace Guidde?
Screen recording — screen recording, walkthroughs and product demos
Exhibit tracking slip
Verdict
Turning a captured click sequence into a narrated video — auto-generating a script describing each step and reading it aloud with local TTS — is a real weekend build combining this catalogue's Scribe-style capture with the TTS approach already described for Murf. What doesn't survive: Guidde's polish in auto-generated narration wording and its team-shared video libraries with analytics.
Exhibit A — The prompt
Received on31.07.2026Build a tool that turns a captured browser workflow into a narrated video — combining this catalogue's Scribe-style click capture with local text-to-speech, same approach as this catalogue's Murf entry. Use a Chrome extension (Manifest V3) to capture a screenshot and a basic templated description at each click, the same mechanism as this catalogue's Scribe entry. After capture, let the user review and edit each step's description text, since auto-generated descriptions are a rough draft, not final copy. Convert each step's finalized description to speech using a local TTS model (Coqui TTS or Piper). Assemble the screenshots and narration into a video using ffmpeg: each screenshot shown for the duration of its narration clip, with a simple zoom-to-hotspot effect on the clicked element. Export as an MP4. Do not build team-shared video libraries with analytics, hosted embeddable players, or multi-language narration — those are out of scope; this produces one narrated video per capture, exported as a file. No API key or account needed; TTS and video assembly run locally.Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 polished auto-generated narration wording
- B.2 team-shared video libraries with analytics
- B.3 video hosting and embeddable players
- B.4 multi-language narration
Prior art
Exhibit C — Why people still pay: capture polish, hosting, and collaboration
Combining screenshots with TTS narration is buildable; writing narration that actually reads naturally, plus hosting and organizing a growing library of these videos for a team, is the ongoing product work.
Questions
Does the narration sound natural?
Reasonably, using an open local TTS model — but the actual wording comes from a basic templated description per step that you're expected to edit before finalizing, not polished auto-generated copy like Guidde's.
Can my team share and track views of these videos?
No — there's no hosting or analytics layer. Each video exports as a standalone MP4 file you'd share manually.
Does it support multiple languages?
Only whatever the local TTS model you choose supports — there's no dedicated multi-language narration feature.
What does it cost to run?
Nothing beyond your own machine — TTS and video assembly run locally with no per-video fee.
Related tools
Receipt