AI Flow: Video

@@ -4,7 +4,7 @@ version = 4  [[package]] name = "adom-aiflow-video"-version = "0.1.1"+version = "0.1.2" dependencies = [  "clap",  "serde",

@@ -2,7 +2,7 @@ # Copied from adom/adom-aiflow template/flow-app. This copy is yours: change anything, never depend on aiflow. [package] name = "adom-aiflow-video"-version = "0.1.1"+version = "0.1.2" edition = "2021" license = "MIT" 

@@ -53,4 +53,4 @@ Every command answers `OK:` with `Hint:` lines. The skill is in [skills/adom-aif  ## Status -0.1.1, first cut. It grew out of the two Ghost Hand films ([john/ghost-hand](https://wiki.adom.inc/john/ghost-hand)). Working today: the flow and its hints, clip recording, the disk estimate, script feedback, the text privacy scan, file tracking and cleanup, and the walkthrough director. Still done by the AI by hand, with the skill as the guide: the cut itself, the frame-by-frame privacy look, the wiki publish and the YouTube upload. AI Flow does not suggest this app at intake yet.+0.1.2, first cut. It grew out of the two Ghost Hand films ([john/ghost-hand](https://wiki.adom.inc/john/ghost-hand)). Working today: the flow and its hints, clip recording, the disk estimate, script feedback, the text privacy scan, file tracking and cleanup, and the walkthrough director. Still done by the AI by hand, with the skill as the guide: the cut itself, the frame-by-frame privacy look, the wiki publish and the YouTube upload. AI Flow does not suggest this app at intake yet.

⋯ 1 unchanged line ⋯

@@ -1,7 +1,7 @@ {   "slug": "adom-aiflow-video",   "type": "app",-  "version": "0.1.1",+  "version": "0.1.2",   "title": "AI Flow: Video",   "description": "Make a narrated walkthrough video of a board built in Adom Hydrogen, put it on the project's wiki page and help the human upload it to YouTube. A dynamic skill: a small Rust binary that suggests each step, records clips, ships a ready-made 3D walkthrough director and gives gentle feedback; the AI does the work. Open source (MIT): the full Rust source ships on this page; build it yourself with ./build.sh.",   "category": "Developer Tools",

@@ -1,7 +1,7 @@ {   "slug": "adom-aiflow-video",   "type": "app",-  "version": "0.1.1",+  "version": "0.1.2",   "title": "AI Flow: Video",   "description": "Make a narrated walkthrough video of a board built in Adom Hydrogen, put it on the project's wiki page and help the human upload it to YouTube. A dynamic skill: a small Rust binary that suggests each step, records clips, ships a ready-made 3D walkthrough director and gives gentle feedback; the AI does the work. Open source (MIT): the full Rust source ships on this page; build it yourself with ./build.sh.",   "category": "Developer Tools",

@@ -64,6 +64,11 @@ Let the human choose, and make either easy. - **AI voiceover**: fine to publish as it is. Say so; do not make them feel it is second best. - **Their own voice**: the biggest step up in quality. Make a private working copy of the draft with the captions showing as a teleprompter, and let them record while they watch, in their own words; they do not have to read it exactly. Then rebuild the film without on-screen captions, fit each scene's footage to what they actually said, and write the captions and chapters again from their recording (check part names and numbers against the script; a transcript gets those wrong). The teleprompter copy is a working file and is never published; published captions stay sidecar files. +  Ask which microphone they want to use; a better mic is worth it:+  - **A recorder of their own** (a wireless lavalier or handheld that records on the device, a field recorder, a phone): they record while watching the teleprompter copy, plug the recorder in or send the file, and you fetch the newest recording. Start the teleprompter with a countdown and ask them to say "start" as it ends, so you can line the recording up; then match each scene by what they said.+  - **The computer's microphone**: record it in sync with the teleprompter copy in a page they open, so the timing is known.+  Either way, listen to the result before you cut: check for clipping, room noise and missed scenes, and offer a retake of single scenes rather than the whole film.+ ## Privacy  A screen recording shows whatever was on the screen. Record only Hydrogen and the build's own windows; a clean second machine is the best protection. `scan` reads the script, captions and description for email addresses, phone numbers, keys and file paths with a user name. Frames need your eyes: look through the contact sheet for mail, chat, names, avatars, other people's windows and toasts. Prefer another take or a tighter crop to a blur, and tell the human what you found.