← PicsartCONTENT HISTORY

Update to Picsart

Snapshot Nov 7, 2026 · 00:09 UTC · version 2.4.0

Collection source: downloaded plugin package.

WHAT CHANGED

Instructions updated for make-screen-videos

Full description — Now says: site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed...

Observed in instructions or declared skills. Runtime behavior has not been tested.

Save Picsart and check future changes in Following.

Up to 3 plugins free · no card required.

All fields and original values.

Full description

Before

Nov 3, 2026 · 18:14 UTC

website into finished videos - several standalone how-to shorts for Reels, TikTok or Shorts that each teach one easy thing, a feature explainer, a promo with each request typed over its result, a timelapse of a long session, or a step-by...

After

Nov 7, 2026 · 00:09 UTC

site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with ...

Skill instructions

Before

Nov 3, 2026 · 18:14 UTC

website into finished videos - several standalone how-to shorts for Reels, TikTok or Shorts that each teach one easy thing, a feature explainer, a promo with each request typed over its result, a timelapse of a long session, or a step-by...

After

Nov 7, 2026 · 00:09 UTC

site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with ...

Supporting files

Before

Nov 3, 2026 · 18:14 UTC

[{"relative_path":"agents/openai.yaml","size_in_bytes":185},{"relative_path":"references/interview.md","size_in_bytes":9652},{"relative_path":"references/outline.md","size_in_bytes":9220},{"relative_path":"references/record.md","size_in_...

After

Nov 7, 2026 · 00:09 UTC

[{"relative_path":"agents/openai.yaml","size_in_bytes":185},{"relative_path":"references/interview.md","size_in_bytes":9893},{"relative_path":"references/outline.md","size_in_bytes":9217},{"relative_path":"references/record.md","size_in_...

Compare saved observations

Download comparison JSON

SKILL.md text changes

− removed · + added

--- before
+++ after
@@ -1,6 +1,6 @@
 ---
 name: make-screen-videos
-description: Turn a screen recording of an app or website into finished videos - several standalone how-to shorts for Reels, TikTok or Shorts that each teach one easy thing, a feature explainer, a promo with each request typed over its result, a timelapse of a long session, or a step-by-step 16:9 tutorial with chapters - in a style the user picks by eye, with a voice-over and captions if they want them. Use when the user has a screen recording and wants a tutorial, a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including adding a generated shot to them.
+description: Turn a screen recording of an app or site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with chapters - in a style picked by eye, with a voice-over and captions if wanted. Use when the user has a screen recording and wants a tutorial video (for a help center too), a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including a generated shot. Not for camera footage or a text-only article.
 metadata:
   uses: screen-recording-analysis, footage-analysis, speech-aware-cutting, screen-recording-layouts, screen-recording-zoom, visual-style, screen-ui-redesign, retention-hooks, scene-building-basics, quality-review, ai-presenter, voice-over-generation, spoken-captions, background-music, motion-design-craft, film-shot-generation
   optional: ai-presenter, voice-over-generation, speech-aware-cutting, spoken-captions, background-music, motion-design-craft, film-shot-generation
@@ -31,6 +31,10 @@
 Asking rules. One job: one or more screen recordings in; one video, or a set of standalone videos,
 out.
 
+Before acting on a stage, load the skills in its row that apply to this job, also when the work resumes there.
+A gate with a scope in brackets opens only there: `SPEND (vertical)` is the cold watch's go-ahead,
+for a vertical video's draft alone; a horizontal video goes straight on through stage 10.
+
 | Stage | Gate | Skills |
 |---|---|---|
 | 1. Take the recording in | — | — |
@@ -39,12 +43,12 @@
 | 4. The set: which videos | SHOW, ASK | screen-recording-analysis, screen-recording-layouts |
 | 5. How we iterate | ASK | — |
 | 6. The style | SHOW, ASK | visual-style, screen-recording-layouts |
-| 7. The plan of video 1 | ASK | retention-hooks, quality-review |
+| 7. The plan of video 1 | ASK | retention-hooks, quality-review, film-shot-generation |
 | 8. The voice, when chosen | ASK, SPEND | voice-over-generation, speech-aware-cutting, spoken-captions, ai-presenter |
-| 9. Build | — | screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style, screen-ui-redesign |
-| 10. Check it | — | quality-review, visual-style, screen-recording-zoom, screen-recording-layouts |
+| 9. Build | — | screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style, screen-ui-redesign, motion-design-craft |
+| 10. Check it | SPEND (vertical) | quality-review, visual-style, screen-recording-zoom, screen-recording-layouts |
 | 11. The user reviews | REVIEW | — |
-| 12. The rest of the set | SPEND, REVIEW | as stages 7-11 |
+| 12. The rest of the set | SPEND, REVIEW | retention-hooks, quality-review, voice-over-generation, screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style |
 | 13. Music, when wanted | ASK, SPEND, REVIEW | background-music |
 | 14. Deliver | DELIVER | — |
 
@@ -90,6 +94,9 @@
 - **Start reading at once** (stage 3): the first question goes in the same turn as the read, so the
   user answers while it runs, and later questions are drafted from what it found ("your recording
   shows five things a viewer could learn").
+- **One question in a turn, never a list.** The only pair is the sound and the ending, as
+  `references/interview.md` marks it. The rest wait for their turn; never preview them. What you
+  settle without asking is said in one line as assumed, so the user can change it.
 - **No plan is written until the message is one sentence.** "You decide" counts: write it from the
   recording and list it as assumed.
 - **The brief closes the interview**: what the user said, what you read, what you assumed, as
@@ -107,7 +114,9 @@
 
 The read is the moments, the still sheets (about four per recording, one sheet of candidate results
 for the set), the stages where they are needed, and the transcript when `input.hasAudio` and the
-voice may be kept. Say in one line that you are reading, and go on; a call that charges goes through
+voice may be kept. It runs no command: `scripts/actions.py` waits for the outline (stage 7). Where
+you cannot run commands, never look for a shell or another way to run a script; read with the
+tools alone, and the build stops as stage 9 says. Say in one line that you are reading, and go on; a call that charges goes through
 picsart-video-rules's SPEND first.
 
 If a read fails twice, tell the user the service is down and to try again shortly. Never guess what
@@ -174,6 +183,10 @@
 outline's words rules say and the retention-hooks skill says. **Never invent what the recording does
 not show.**
 
+With the plan, write its viewer track, as quality-review's cold watch says: the promise by 3 s,
+each step a newcomer must follow, and the payoff, with the questions to ask after watching. Its
+questions are shown with the rows, for correction.
+
 **Review the script before anyone sees the plan**, as quality-review's script review says: put its
 rewritten lines into the rows, and its why under the opening sentence. A `drop` is not planned:
 ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what to do with that video, with the reason quoted: "Leave it out" (recommended: the
@@ -185,12 +198,12 @@
 skill's hook rules say, each saying what it opens on. The rows are shown for correction, not for
 approval: anything the user writes changes them, and their answer to the hook moves on.
 
-**The viewer panel, before the ASK** (a vertical video). Build each drafted hook, which is free,
-and run quality-review's viewer panel on its first 2 s. Rewrite a hook that fails, as the panel's
-`fix` says. Offer only hooks that pass, each with the panel's one line. This comes before
-anything is paid for: a hook that opens on footage not made yet is judged on its still or its
-description. When no hook passes after the rounds, offer the best one and say what still
-fails.
+**Each hook built, before the ASK** (a vertical video). Build each drafted hook, which is free,
+take its first 2 s on a contact sheet, and check it against the retention-hooks skill's hook rules.
+Rewrite one that breaks them. Offer each as a prediction, never as tested: nobody has watched it
+yet. A hook that opens on footage not made yet is judged on its still or its description. The
+draft's cold watch is the hook's test: its first two pieces (to 3 s) say whether strangers keep
+watching.
 
 An opening the recording does not hold, such as a generated shot, is a SPEND. **The first reply
 about a generated shot offers its key frames and nothing else:** stills at the hook, the turn and
@@ -231,7 +244,7 @@
   probe each line's real length, and transcribe each for its captions, as the voice-over-generation and
   spoken-captions skills say. The line sets the beat's length; the picture holds.
 - **The user's own voice** (`input.hasAudio`, a transcript that is not thin): kept whole, never cut
-  mid-sentence, captioned from the saved transcript.
+  mid-sentence, captioned from its transcript as `picsart_media_transcribe` returned it.
 - **A presenter on camera**: the promo route only, never the default, as `references/routes.md`
   says.
 - **Captions and music, or text only**: nothing to generate here.
@@ -252,8 +265,9 @@
 3. If it exits with an error instead, its one line names the field it cannot build from. Fix that
    field and run it again.
 
-Building is free and calls no tool. **Where you cannot run Python**, say that the video cannot be
-built here, and stop.
+Building is free and calls no tool. **Where you cannot run Python**, the build stops as
+picsart-video-rules says for a script that builds the scene: say this video needs the scene builder,
+offer to carry on on the desktop or in the Picsart app from the record, and stop. Never hand-build it.
 
 A timelapse omits stages with no usable footage inside their source bounds and cut guards.
 
@@ -300,10 +314,12 @@
   with each step's measured reveal as a change);
 - the first frame is a real moment (the payoff, or the problem it solves), not a login screen or
   an empty app;
-- a vertical video's draft passes quality-review's viewer panel. Its hook is judged again on the real
-  frames; its `summary` goes in the review line.
+- a vertical video's draft gets quality-review's cold watch against its viewer track, quoted first
+  as it says, with its SPEND (quote it in one line, then get the yes as an ASK with "Yes, go ahead" / "Yes, up to N credits for this task" / "No": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit ("as much as needed"), never inferred from words like "go ahead". Within it, ask nothing: quote it in one line ("About 8 credits; 12 of 20 left after this"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) before round 1 unless every call is quoted at 0 credits: its first
+  two pieces test the hook, and its floor line goes in the review line. When the user says no, the
+  review line says "no cold watch yet";
 - then quality-review's critic panel runs on every draft (the six `critic-*` agents where the host has
-  them), with the recording's stills at each step as its source. Make its must-fixes, and ASK its `choices` before REVIEW, never applied silently.
+  them), with the recording's stills at each step as its source and the cold watch's gap list. Make its must-fixes, and ASK its `choices` before REVIEW, never applied silently.
 
 Then grade the brief with quality-review's `checklist.py`: the message, the destinations, the length,
 the hook and the ending from the record, and the plan's step lines. Fix only what it says no to.
@@ -327,9 +343,11 @@
 
 ## 11. The user reviews
 
-REVIEW the mask build, titled "<Product> <kind> — draft 1", with the checklist's `summary` and one
-line on what to look at for this pass, as `references/routes.md` gives it ("Does the result land in the first two seconds? Is
-each click readable? Say where it drags"). Each new draft's number goes up by one. Comments apply
+REVIEW the mask build, titled "<Product> <kind> — draft 1", opened cold as picsart-video-rules
+says: its length and the track's questions. Once the user has answered, give the checklist's
+`summary`, the cold watch's floor line and the line `references/routes.md` gives for the kind ("Does
+the result land in the first two seconds? Is each click readable? Say where it drags"). Each new
+draft's number goes up by one. Comments apply
 where they point; after the third round on the same video, ask what is causing the misses (the
 look, the pace or the words) instead of patching again.
 
Full technical diff · 3 changed fields

changed /description

BEFORE
"Turn a screen recording of an app or website into finished videos - several standalone how-to shorts for Reels, TikTok or Shorts that each teach one easy thing, a feature explainer, a promo with each request typed over its result, a timelapse of a long session, or a step-by-step 16:9 tutorial with chapters - in a style the user picks by eye, with a voice-over and captions if they want them. Use when the user has a screen recording and wants a tutorial, a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including adding a generated shot to them."
AFTER
"Turn a screen recording of an app or site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with chapters - in a style picked by eye, with a voice-over and captions if wanted. Use when the user has a screen recording and wants a tutorial video (for a help center too), a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including a generated shot. Not for camera footage or a text-only article."

changed /included_files

BEFORE
[
  {
    "relative_path": "agents/openai.yaml",
    "size_in_bytes": 185
  },
  {
    "relative_path": "references/interview.md",
    "size_in_bytes": 9652
  },
  {
    "relative_path": "references/outline.md",
    "size_in_bytes": 9220
  },
  {
    "relative_path": "references/record.md",
    "size_in_bytes": 1849
  },
  {
    "relative_path": "references/routes.md",
    "size_in_bytes": 12432
  },
  {
    "relative_path": "references/script.md",
    "size_in_bytes": 2023
  },
  {
    "relative_path": "references/spec.md",
    "size_in_bytes": 15214
  },
  {
    "relative_path": "scripts/_build.py",
    "size_in_bytes": 33482
  },
  {
    "relative_path": "scripts/_diagnostics.py",
    "size_in_bytes": 21558
  },
  {
    "relative_path": "scripts/_geometry.py",
    "size_in_bytes": 1139
  },
  {
    "relative_path": "scripts/_input.py",
    "size_in_bytes": 12663
  },
  {
    "relative_path": "scripts/_js.py",
    "size_in_bytes": 4625
  },
  {
    "relative_path": "scripts/_moments.py",
    "size_in_bytes": 1497
  },
  {
    "relative_path": "scripts/_preview.py",
    "size_in_bytes": 5743
  },
  {
    "relative_path": "scripts/_scene.py",
    "size_in_bytes": 21419
  },
  {
    "relative_path": "scripts/_stages.py",
    "size_in_bytes": 5261
  },
  {
    "relative_path": "scripts/_steps.py",
    "size_in_bytes": 1687
  },
  {
    "relative_path": "scripts/_timeedit.py",
    "size_in_bytes": 43226
  },
  {
    "relative_path": "scripts/_tutorial_edit.py",
    "size_in_bytes": 14003
  },
  {
    "relative_path": "scripts/_tutorial_input.py",
    "size_in_bytes": 9273
  },
  {
    "relative_path": "scripts/_tutorial_layout.py",
    "size_in_bytes": 12483
  },
  {
    "relative_path": "scripts/actions.py",
    "size_in_bytes": 11047
  },
  {
    "relative_path": "scripts/adherence.py",
    "size_in_bytes": 10597
  },
  {
    "relative_path": "scripts/cover.py",
    "size_in_bytes": 16819
  },
  {
    "relative_path": "scripts/framing.py",
    "size_in_bytes": 7644
  },
  {
    "relative_path": "scripts/panels.py",
    "size_in_bytes": 13591
  },
  {
    "relative_path": "scripts/private_text.py",
    "size_in_bytes": 17020
  },
  {
    "relative_path": "scripts/promo_scene.py",
    "size_in_bytes": 2889
  },
  {
    "relative_path": "scripts/readability.py",
    "size_in_bytes": 7792
  },
  {
    "relative_path": "scripts/safe_zone.py",
    "size_in_bytes": 61049
  },
  {
    "relative_path": "scripts/style.py",
    "size_in_bytes": 7576
  },
  {
    "relative_path": "scripts/tutorial_scene.py",
    "size_in_bytes": 17091
  },
  {
    "relative_path": "scripts/zoom.py",
    "size_in_bytes": 15530
  },
  {
    "relative_path": "scripts/zoom_plan.py",
    "size_in_bytes": 16698
  }
]
AFTER
[
  {
    "relative_path": "agents/openai.yaml",
    "size_in_bytes": 185
  },
  {
    "relative_path": "references/interview.md",
    "size_in_bytes": 9893
  },
  {
    "relative_path": "references/outline.md",
    "size_in_bytes": 9217
  },
  {
    "relative_path": "references/record.md",
    "size_in_bytes": 1849
  },
  {
    "relative_path": "references/routes.md",
    "size_in_bytes": 12444
  },
  {
    "relative_path": "references/script.md",
    "size_in_bytes": 2023
  },
  {
    "relative_path": "references/spec.md",
    "size_in_bytes": 15214
  },
  {
    "relative_path": "scripts/_build.py",
    "size_in_bytes": 33482
  },
  {
    "relative_path": "scripts/_diagnostics.py",
    "size_in_bytes": 21558
  },
  {
    "relative_path": "scripts/_geometry.py",
    "size_in_bytes": 1139
  },
  {
    "relative_path": "scripts/_input.py",
    "size_in_bytes": 12663
  },
  {
    "relative_path": "scripts/_js.py",
    "size_in_bytes": 4625
  },
  {
    "relative_path": "scripts/_moments.py",
    "size_in_bytes": 1497
  },
  {
    "relative_path": "scripts/_preview.py",
    "size_in_bytes": 5743
  },
  {
    "relative_path": "scripts/_scene.py",
    "size_in_bytes": 21419
  },
  {
    "relative_path": "scripts/_stages.py",
    "size_in_bytes": 5261
  },
  {
    "relative_path": "scripts/_steps.py",
    "size_in_bytes": 1687
  },
  {
    "relative_path": "scripts/_timeedit.py",
    "size_in_bytes": 43226
  },
  {
    "relative_path": "scripts/_tutorial_edit.py",
    "size_in_bytes": 14003
  },
  {
    "relative_path": "scripts/_tutorial_input.py",
    "size_in_bytes": 9273
  },
  {
    "relative_path": "scripts/_tutorial_layout.py",
    "size_in_bytes": 12483
  },
  {
    "relative_path": "scripts/actions.py",
    "size_in_bytes": 11047
  },
  {
    "relative_path": "scripts/adherence.py",
    "size_in_bytes": 10597
  },
  {
    "relative_path": "scripts/cover.py",
    "size_in_bytes": 16819
  },
  {
    "relative_path": "scripts/framing.py",
    "size_in_bytes": 7644
  },
  {
    "relative_path": "scripts/panels.py",
    "size_in_bytes": 13591
  },
  {
    "relative_path": "scripts/private_text.py",
    "size_in_bytes": 17020
  },
  {
    "relative_path": "scripts/promo_scene.py",
    "size_in_bytes": 2889
  },
  {
    "relative_path": "scripts/readability.py",
    "size_in_bytes": 7792
  },
  {
    "relative_path": "scripts/safe_zone.py",
    "size_in_bytes": 61049
  },
  {
    "relative_path": "scripts/style.py",
    "size_in_bytes": 7576
  },
  {
    "relative_path": "scripts/tutorial_scene.py",
    "size_in_bytes": 17091
  },
  {
    "relative_path": "scripts/zoom.py",
    "size_in_bytes": 15530
  },
  {
    "relative_path": "scripts/zoom_plan.py",
    "size_in_bytes": 16698
  }
]

changed /skill_md_contents

BEFORE
"---\nname: make-screen-videos\ndescription: Turn a screen recording of an app or website into finished videos - several standalone how-to shorts for Reels, TikTok or Shorts that each teach one easy thing, a feature explainer, a promo with each request typed over its result, a timelapse of a long session, or a step-by-step 16:9 tutorial with chapters - in a style the user picks by eye, with a voice-over and captions if they want them. Use when the user has a screen recording and wants a tutorial, a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including adding a generated shot to them.\nmetadata:\n  uses: screen-recording-analysis, footage-analysis, speech-aware-cutting, screen-recording-layouts, screen-recording-zoom, visual-style, screen-ui-redesign, retention-hooks, scene-building-basics, quality-review, ai-presenter, voice-over-generation, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  optional: ai-presenter, voice-over-generation, speech-aware-cutting, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  imports: screen-recording-analysis, screen-recording-layouts, screen-recording-zoom, visual-style\n  title: Make screen videos\n  short-description: Turn a screen recording into how-to videos\n  argument-hint: '[screen recording and the video you want]'\n  starter-prompt: Turn my screen recording into short how-to videos.\n---\n\n# Screen video\n\nA screen recording is mostly waiting, and what it shows is a few pixels of dense UI. The same\nrecording can become very different videos: a set of 20-second shorts that each show how easy one\nthing is, a promo that sells, or a tutorial that keeps every step. Which one is the user's to say,\nso this use case starts with a short interview, never with a draft to approve.\n\nThe kinds, each a route in `references/routes.md`:\n\n| Kind | For | Frame | Built by |\n|---|---|---|---|\n| **How-to short** | one easy task, start to result: \"look how easy\" | 9:16 | `scripts/promo_scene.py` |\n| **Feature explainer** | one feature: the problem, the feature, two or three uses | 9:16 or 16:9 | `scripts/promo_scene.py` |\n| **Promo** | the best results, fast, to sell; a timelapse for a long session; a re-cut for an edited video | 9:16 | `scripts/promo_scene.py` |\n| **Tutorial** | every step in order, held long enough to follow, with chapters | 16:9 | `scripts/tutorial_scene.py` |\n\nRead picsart-video-rules first: how to talk and how to ask are there, and every question here follows its\nAsking rules. One job: one or more screen recordings in; one video, or a set of standalone videos,\nout.\n\n| Stage | Gate | Skills |\n|---|---|---|\n| 1. Take the recording in | — | — |\n| 2. The interview | ASK | — |\n| 3. Read the recording | — | screen-recording-analysis, footage-analysis |\n| 4. The set: which videos | SHOW, ASK | screen-recording-analysis, screen-recording-layouts |\n| 5. How we iterate | ASK | — |\n| 6. The style | SHOW, ASK | visual-style, screen-recording-layouts |\n| 7. The plan of video 1 | ASK | retention-hooks, quality-review |\n| 8. The voice, when chosen | ASK, SPEND | voice-over-generation, speech-aware-cutting, spoken-captions, ai-presenter |\n| 9. Build | — | screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style, screen-ui-redesign |\n| 10. Check it | — | quality-review, visual-style, screen-recording-zoom, screen-recording-layouts |\n| 11. The user reviews | REVIEW | — |\n| 12. The rest of the set | SPEND, REVIEW | as stages 7-11 |\n| 13. Music, when wanted | ASK, SPEND, REVIEW | background-music |\n| 14. Deliver | DELIVER | — |\n\n| Read this | Before you |\n|---|---|\n| `references/interview.md` | ask the first question, and at every ASK in stages 2-7 |\n| `references/routes.md` | write a plan, build, check or deliver: what differs by kind |\n| `references/spec.md` | write a vertical video's spec or build it |\n| `references/outline.md` | write a tutorial's outline or build it |\n| `references/script.md` | write a presenter's lines |\n| `references/record.md` | write or resume the project record |\n\n**The challenger at each decision.** Before you go on from a decision, quality-review's challenger (the\n`challenger` agent where the host has it) asks \"why this?\" of it: the set of videos (stage 4), the\nstyle frames (6), the plan and its hooks before the ASK (7), the first build's UI, against the\nrecording's own screens (9), and every quote before its SPEND (8, 12, 13). Act on its verdict as\nquality-review says. A `stop`, or a `rethink` of the user's own choice, goes to the user in the next ASK,\nthe cheaper alternative first, before anything is spent.\n\n## 1. Take the recording in\n\n- **Read the user's brief first.** Every file in the material that is not media, read in full\n  before the first question, as picsart-video-rules's The user's own brief says. A brief sets the message, the\n  audience and the must-shows.\n- **No recording yet → open the upload in your first reply.** When the message carries no recording\n  (no attachment, no link, no Drive file named), call\n  `picsart_drive_upload_ask({ accept: 'video', purpose: 'your screen recording' })` in that same turn,\n  with one line on what happens once it is in, and ask the interview's first question beside it.\n  If the host cannot show the upload, follow picsart-video-rules's host notes.\n- A link the user pasted, or a file in their Picsart Drive: bring it in as picsart-video-rules's Asset URLs\n  say. Every tool below reads the recording by the URL that comes back.\n- **Several recordings** are one set of material: read each, and topics may come from any of them.\n  One video plays one recording.\n\n## 2. The interview\n\nAsk as `references/interview.md` says: one question per turn, in its order, each skipped when the\nuser's words or the recording already answer it. A typical run asks four to six. The questions are\nwhere it goes, what a viewer should get from it, how many videos, who it is for, the message, the\nlength (only when the destination leaves a range), the sound and the ending, and whether the\naccount on screen is real.\n\n- **Start reading at once** (stage 3): the first question goes in the same turn as the read, so the\n  user answers while it runs, and later questions are drafted from what it found (\"your recording\n  shows five things a viewer could learn\").\n- **No plan is written until the message is one sentence.** \"You decide\" counts: write it from the\n  recording and list it as assumed.\n- **The brief closes the interview**: what the user said, what you read, what you assumed, as\n  picsart-video-rules says, with how to change each assumption. The confirmation is the next real decision\n  (stage 4), never a yes.\n\nThe user can say \"just make it\" at any point: take the recommended answers, list them as assumed,\nand go on to the next choice made by eye or the next charge.\n\n## 3. Read the recording\n\nRead it as the screen-recording-analysis skill says: the moments first, then still sheets of the moments that\nmatter, and for a long session its stages. The recording is never sent to a model that watches\nvideo: never call `picsart_media_describe_video`. footage-analysis is used only for the transcript.\n\nThe read is the moments, the still sheets (about four per recording, one sheet of candidate results\nfor the set), the stages where they are needed, and the transcript when `input.hasAudio` and the\nvoice may be kept. Say in one line that you are reading, and go on; a call that charges goes through\npicsart-video-rules's SPEND first.\n\nIf a read fails twice, tell the user the service is down and to try again shortly. Never guess what\nthe recording shows. **Moments already read in this session are not read again.**\n\nWhat the read decides, said in one line each and never asked:\n\n- **The shape.** `input.orientation` sets the layout, as the screen-recording-layouts skill says: a portrait\n  recording fills the frame, a square one fills its width. A tutorial needs a raw landscape\n  recording; a portrait or edited one gets the vertical kinds instead (`references/routes.md`).\n- **The tier**, for a promo: prompt over result, timelapse or re-cut, from `tier` and `tierReason`,\n  checked on the first sheet. When the sheet disagrees, believe the sheet.\n- **What the recording is:** the product, what its steps get done, its own colours (for the\n  style's brand), and whether there is speech.\n\n## 4. The set: which videos\n\nFind the candidates as `references/routes.md` says for the kind: standalone topics for how-to\nshorts (each a run of actions from a state a viewer can reach to a visible result), the features\nfor an explainer, the steps that sell for a promo. **Every candidate comes from a still you read.**\n\nA result that is only words in a chat reply, or that comes after the recording ends, is not on a\nstill: leave that candidate out, and say in one line what footage would bring it back.\n\nSHOW one sheet of the candidates' results, then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which to make, as `references/interview.md`\nsays: a multi-select, the most teachable first, each with what the viewer learns and where it is\nin the recording. A tutorial skips this stage: it keeps every step, and its outline is stage 7.\n\nEvery short stands alone: its own hook, its own payoff, its own ending. Never \"part 2\", never a\nreference to another short.\n\n**Will the screens read?** For the vertical kinds, once the videos are chosen, measure each moment\nthey will show with the screen-recording-layouts skill's `readability.py`: its box, the smallest text the viewer\nmust read and the words in it (from the still sheets), its event's luma, and the recommended\nstyle's canvas. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) \"The screens\" (`references/interview.md`), its own question, with the\nmeasured reason: when a moment fails, or when the kind is a how-to or a promo where the UI is only\nthe means. A tutorial keeps the recording: it teaches where to click in the real app.\n\n## 5. How we iterate\n\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what the user wants to see first (`references/interview.md`): stills of video 1 in three\nstyles, video 1 finished, or the written plan of every video. Then say the agreed plan in one line,\nas picsart-video-rules's Asking says, with what costs: \"Then: you pick a style on three stills, I build\nshort 1, you review it (free, as many rounds as you like), then I build the other four in the same\nstyle and you review them together. Only the voice costs, quoted first.\" Store it in the record\nand repeat its line at each pass.\n\n## 6. The style\n\nChoose it as the visual-style skill says. Recommend one from the recording and the brief, with its\nreason; build video 1's hook frame and one step frame in it and in two styles that differ, with the\nrecording's colours as the brand where they fit; and SHOW them as three contact sheets on the user's\nown recording. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which style, the recommended one first. Building and the contact sheet are\nfree. A mix (\"the colours of one, the type of the other\") comes through Other: build it and show\nit.\n\nThe pick is the set's style. Save it in the record; every video of the set uses it.\n\n## 7. The plan of video 1\n\nWrite the plan as `references/routes.md` says for the kind. It opens with one sentence, \"This\nshort shows <audience> that <message>\", and has one row per beat: the time, what is on screen, the\nwords, and why the beat is there, traced to the message. Write the words as the spec's or the\noutline's words rules say and the retention-hooks skill says. **Never invent what the recording does\nnot show.**\n\n**Review the script before anyone sees the plan**, as quality-review's script review says: put its\nrewritten lines into the rows, and its why under the opening sentence. A `drop` is not planned:\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what to do with that video, with the reason quoted: \"Leave it out\" (recommended: the\nrecording never shows its result) / \"Teach the smaller lesson\" (the one the review names) / \"I'll\nrecord the result\" (the footage its `rescue` names). The review is free, so it never asks to run.\n\nThe ASK is the hook: 2-3 drafted openings that differ in what they open on (a problem the viewer\nfeels, the result in the middle of happening, a before and after), each written as the retention-hooks\nskill's hook rules say, each saying what it opens on. The rows are shown for correction, not for\napproval: anything the user writes changes them, and their answer to the hook moves on.\n\n**The viewer panel, before the ASK** (a vertical video). Build each drafted hook, which is free,\nand run quality-review's viewer panel on its first 2 s. Rewrite a hook that fails, as the panel's\n`fix` says. Offer only hooks that pass, each with the panel's one line. This comes before\nanything is paid for: a hook that opens on footage not made yet is judged on its still or its\ndescription. When no hook passes after the rounds, offer the best one and say what still\nfails.\n\nAn opening the recording does not hold, such as a generated shot, is a SPEND. **The first reply\nabout a generated shot offers its key frames and nothing else:** stills at the hook, the turn and\nthe payoff, quoted on their own, shown together before any video is priced, and reused to open the\nshot. No video model is read, compiled or priced until the user approves the stills. **Pictures the\nuser brings, a film's scene pictures included, are references for those stills, never the stills\nthemselves**: they are not this short's frame or its beats, so the key frames are still drafted\nfrom them, quoted and shown first, even when the user says to go ahead. That reply says all four\nto the user, in its own words, beside the quote:\n\n- **Every key point**: a still for each generated beat (a generated hook and payoff are two stills,\n  one batch, one quote), and the recording's own frame at the turn, which is free.\n- **The quote covers those stills alone.** No video price, take or model in it.\n- **They are shown together** as the concept before any video is priced.\n- **Each approved still opens its shot** as its start frame; none is a throwaway preview.\n\nFor example: \"You'll see the hook and payoff stills together, beside the recording's frame at the\nturn, before any video is priced; each one you approve becomes the first frame of its shot.\"\n\nAn approved plan already names its hook, turn and payoff, so this reply runs no script: take the\nkey points from the plan, and never search for a shell or hand the script to another agent. Draft it first and\nquote the draft beside the final, as picsart-video-rules's SPEND says; a crop of the recording is usually\nsmaller than the frame, so say so when it is the reference.\nIts draft is key frames, as picsart-video-rules's SPEND rule 8 says: plan them on this plan's beats with the\nfilm-shot-generation skill's `scripts/keyframes.py`, SHOW them before the shot is quoted, and open the shot on\nthe approved one. In the hook ASK, a generated opening is offered as its key frames first.\nIts key frames and its shot are compiled with `picsart_film_compile_prompt` (picsart-video-rules's SPEND rule 9), with\nthe size and lens the recording's own plan gives that shot.\n\nThe words' language is the user's own, unless the recording's words are in another: then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which.\n\n## 8. The voice, when chosen\n\nThe sound was chosen in the interview. Make it as `references/routes.md` says:\n\n- **A generated voice**: one line per beat, quoted as one batch with `picsart_preflight` after\n  reading the voice model with `picsart_model_params` (picsart-video-rules, SPEND). Generate after the yes,\n  probe each line's real length, and transcribe each for its captions, as the voice-over-generation and\n  spoken-captions skills say. The line sets the beat's length; the picture holds.\n- **The user's own voice** (`input.hasAudio`, a transcript that is not thin): kept whole, never cut\n  mid-sentence, captioned from the saved transcript.\n- **A presenter on camera**: the promo route only, never the default, as `references/routes.md`\n  says.\n- **Captions and music, or text only**: nothing to generate here.\n\nSeveral videos' voice lines are one batch: one quote and one yes for the set.\n\n## 9. Build\n\nBuild with the script for the kind, from this use case's folder. Run it; never work the scene out\nyourself.\n\n1. Write its input to a file as `references/spec.md` (vertical kinds) or `references/outline.md`\n   (tutorial) says, with `style` and `brand` from stage 6 in the spec, `clip: \"mask\"`, and\n   `moments` exactly as `picsart_media_screen_moments` returned them.\n2. Run `python3 scripts/promo_scene.py <file>` or `python3 scripts/tutorial_scene.py <file>`. It\n   prints one JSON line with the scene, its length, a timeline and diagnostics. The skills'\n   scripts it imports ship beside it in `scripts/`.\n3. If it exits with an error instead, its one line names the field it cannot build from. Fix that\n   field and run it again.\n\nBuilding is free and calls no tool. **Where you cannot run Python**, say that the video cannot be\nbuilt here, and stop.\n\nA timelapse omits stages with no usable footage inside their source bounds and cut guards.\n\n**Act on every diagnostic before you look at anything else.** An `error` means the scene is wrong:\nmake the change its `fix` names and build again; never patch the scene past one. A `warning` means\nit plays but may read wrong: fix it when the fix keeps what the user chose, otherwise look at that\nmoment at the check and say it in the review line. An `error` still there after two builds: tell\nthe user what it says and what you changed, and stop there.\n\n**A step to recreate** (the user chose designed UI for it): after the build, write the step's\nelements and states from the still notes as the screen-ui-redesign skill says, with the step's panel as\n`box` and its timeline span as `start` and `duration`, and run its `scripts/ui_beat.py`. With\n`picsart_scene_patch`, remove the step's `<id>_panel` and `<id>_rim` and add the beat's\nlayers and assets in their place. Do it again after every build, as the covers are. The redraw is\nthe app's own screen, larger and cleaner, never an invented one: before the video is rendered,\ncompare each state's frame with its `source` still, as the screen-ui-redesign skill says.\n\nCaptions, a presenter's takes and callouts go on after the build with\n`picsart_scene_patch`, as `references/routes.md` says. The build's pushes are its own; a\npush you add by hand, on a scene the user edited, comes from the screen-recording-zoom skill's\n`zoom_plan.py`, `balanced` unless the user asked for gentler or punchier pushes. When the user's comments change what a\nvideo shows or says, change the spec or outline and build again; when they edited the scene in the\neditor, their document replaces yours (picsart-video-rules): patch it instead.\n\n**A generated shot goes in as a presenter's take does**: build with `hook.duration` set to its\nlength, then patch it in full frame with `picsart_scene_patch` (`references/routes.md`).\nNever swap the script for a scene of your own. A scene worked out by hand skips the build's\ndiagnostics, and the stages around it still hold: key frames and SPEND before, the check and REVIEW\nafter.\n\n## 10. Check it\n\nValidate the build with `picsart_scene_validate`, then check it with quality-review, at most 8\nmoments to a contact sheet: the hook, the end of each beat where its push has landed, the ending,\nand the moment of each warning left in. Ask the questions `references/routes.md` gives for the\nkind. Then the polish pass, which quality-review runs:\n\n- the build's colours and entrances keep to its style (visual-style's `adherence.py`);\n- nothing on the visual-style skill's list of defaults that look machine-made, unless it is\n  kept with a reason;\n- every word and the call to action sit inside the destination's safe zone: the screen-recording-layouts skill's\n  `safe_zone.py`, with the profile of each place in the brief;\n- every line is held long enough to read and the picture never stalls (quality-review's `timing.py`,\n  with each step's measured reveal as a change);\n- the first frame is a real moment (the payoff, or the problem it solves), not a login screen or\n  an empty app;\n- a vertical video's draft passes quality-review's viewer panel. Its hook is judged again on the real\n  frames; its `summary` goes in the review line.\n- then quality-review's critic panel runs on every draft (the six `critic-*` agents where the host has\n  them), with the recording's stills at each step as its source. Make its must-fixes, and ASK its `choices` before REVIEW, never applied silently.\n\nThen grade the brief with quality-review's `checklist.py`: the message, the destinations, the length,\nthe hook and the ending from the record, and the plan's step lines. Fix only what it says no to.\n\n**Private data, before anything is shown.** Run the screen-recording-analysis skill's privacy pass on every\ntext you read for this video, on its stills and in its transcript. With real data, cover each\nhit it keeps, and each name or address you saw, with the screen-recording-layouts skill's `cover.py`, in the\nstyle's surface colour, and patch the cards in; with a demo account, the keys, tokens and card\nnumbers only. A box on screen for a whole step can be cut or framed out instead. Cover again after\nevery build: a build drops the patched cards. Never open REVIEW with a hit uncovered; say in the\nreview line what was covered.\n\n**Masks first.** The first contact sheet must include a panel. If it fails with `the authored mask\ncarries no clip`, masks are not rendering here (the screen-recording-layouts skill has the dated fact): build the\nsame spec again with `clip: \"nest\"` and check that build. REVIEW opens the mask build; the nest\nbuild is the one that renders.\n\n**Judge legibility on a still, not on softness alone**: crop a `picsart_media_video_stills` sheet\nto the beat's box at that moment. If the still is sharp, the recording is legible; never push\nharder to fix a soft contact sheet (screen-recording-zoom).\n\n## 11. The user reviews\n\nREVIEW the mask build, titled \"<Product> <kind> — draft 1\", with the checklist's `summary` and one\nline on what to look at for this pass, as `references/routes.md` gives it (\"Does the result land in the first two seconds? Is\neach click readable? Say where it drags\"). Each new draft's number goes up by one. Comments apply\nwhere they point; after the third round on the same video, ask what is causing the misses (the\nlook, the pace or the words) instead of patching again.\n\n## 12. The rest of the set\n\nBuild the other videos with video 1's approved style and with what the user's comments on it\ntaught (the pace, the words' tone, the moves). Plan and review each as stage 7 does, with every `drop` in the set asked in one turn, and its own hook asked\nonly when the user chose to see plans; otherwise draft the hook yourself and list it as assumed.\nTheir voice lines are one batch SPEND. REVIEW them one after another, and ask one \"anything to\nchange?\" for the set (picsart-video-rules).\n\n## 13. Music, when wanted\n\nOffer it once, after the drafts are approved, unless the sound chosen was the user's own voice or\nthe recording's own sound. ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) with two drafted briefs that differ in feel, as the background-music skill\nsays, and \"No music\". On a pick, the bed (and a promo's push sounds, `references/routes.md`) is one\nSPEND (quote it in one line, then get the yes as an ASK with \"Yes, go ahead\" / \"Yes, up to N credits for this task\" / \"No\": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit (\"as much as needed\"), never inferred from words like \"go ahead\". Within it, ask nothing: quote it in one line (\"About 8 credits; 12 of 20 left after this\"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) for the set. Lay it under each video as the background-music skill says, validate and look as stage\n10 does, and REVIEW, titled \"<Product> <kind> — music\", with one line on what to listen for. If it\nfails or is declined, say so in one line and go on with the approved drafts.\n\n## 14. Deliver\n\nDELIVER each approved video: the mask build if masks rendered at the check, otherwise the nest\nbuild. Check each export against its scene with quality-review's `qc.py` before you give its link. If the user edited a scene in the editor and masks do not render, export their scene; if\nthat fails on its masks, say so and offer the nest build, naming the edits it leaves out.\n\nWith each video, give:\n\n- the link, its length against the target, and its status: done, done with concerns (each named),\n  or blocked;\n- **post copy** for the destination, as `references/routes.md` says: a title under 40 characters,\n  a first caption line that hooks and a caption that says what the video shows, and 3-5 words of\n  cover text taken from the payoff frame; hashtags only if asked;\n- what it sounds like: silent (a track can be added when it is posted), the user's voice, a\n  generated voice (AI-made: tell them to switch on the platform's AI-content label), or music;\n- a tutorial's chapters, exactly as the build printed them, to paste into the description.\n\nThen offer one next step in a sentence: another kind from the same recording (nothing is read\nagain), or saving this set's brief and style in the record so \"make another\" starts from them.\n"
AFTER
"---\nname: make-screen-videos\ndescription: Turn a screen recording of an app or site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with chapters - in a style picked by eye, with a voice-over and captions if wanted. Use when the user has a screen recording and wants a tutorial video (for a help center too), a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including a generated shot. Not for camera footage or a text-only article.\nmetadata:\n  uses: screen-recording-analysis, footage-analysis, speech-aware-cutting, screen-recording-layouts, screen-recording-zoom, visual-style, screen-ui-redesign, retention-hooks, scene-building-basics, quality-review, ai-presenter, voice-over-generation, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  optional: ai-presenter, voice-over-generation, speech-aware-cutting, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  imports: screen-recording-analysis, screen-recording-layouts, screen-recording-zoom, visual-style\n  title: Make screen videos\n  short-description: Turn a screen recording into how-to videos\n  argument-hint: '[screen recording and the video you want]'\n  starter-prompt: Turn my screen recording into short how-to videos.\n---\n\n# Screen video\n\nA screen recording is mostly waiting, and what it shows is a few pixels of dense UI. The same\nrecording can become very different videos: a set of 20-second shorts that each show how easy one\nthing is, a promo that sells, or a tutorial that keeps every step. Which one is the user's to say,\nso this use case starts with a short interview, never with a draft to approve.\n\nThe kinds, each a route in `references/routes.md`:\n\n| Kind | For | Frame | Built by |\n|---|---|---|---|\n| **How-to short** | one easy task, start to result: \"look how easy\" | 9:16 | `scripts/promo_scene.py` |\n| **Feature explainer** | one feature: the problem, the feature, two or three uses | 9:16 or 16:9 | `scripts/promo_scene.py` |\n| **Promo** | the best results, fast, to sell; a timelapse for a long session; a re-cut for an edited video | 9:16 | `scripts/promo_scene.py` |\n| **Tutorial** | every step in order, held long enough to follow, with chapters | 16:9 | `scripts/tutorial_scene.py` |\n\nRead picsart-video-rules first: how to talk and how to ask are there, and every question here follows its\nAsking rules. One job: one or more screen recordings in; one video, or a set of standalone videos,\nout.\n\nBefore acting on a stage, load the skills in its row that apply to this job, also when the work resumes there.\nA gate with a scope in brackets opens only there: `SPEND (vertical)` is the cold watch's go-ahead,\nfor a vertical video's draft alone; a horizontal video goes straight on through stage 10.\n\n| Stage | Gate | Skills |\n|---|---|---|\n| 1. Take the recording in | — | — |\n| 2. The interview | ASK | — |\n| 3. Read the recording | — | screen-recording-analysis, footage-analysis |\n| 4. The set: which videos | SHOW, ASK | screen-recording-analysis, screen-recording-layouts |\n| 5. How we iterate | ASK | — |\n| 6. The style | SHOW, ASK | visual-style, screen-recording-layouts |\n| 7. The plan of video 1 | ASK | retention-hooks, quality-review, film-shot-generation |\n| 8. The voice, when chosen | ASK, SPEND | voice-over-generation, speech-aware-cutting, spoken-captions, ai-presenter |\n| 9. Build | — | screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style, screen-ui-redesign, motion-design-craft |\n| 10. Check it | SPEND (vertical) | quality-review, visual-style, screen-recording-zoom, screen-recording-layouts |\n| 11. The user reviews | REVIEW | — |\n| 12. The rest of the set | SPEND, REVIEW | retention-hooks, quality-review, voice-over-generation, screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style |\n| 13. Music, when wanted | ASK, SPEND, REVIEW | background-music |\n| 14. Deliver | DELIVER | — |\n\n| Read this | Before you |\n|---|---|\n| `references/interview.md` | ask the first question, and at every ASK in stages 2-7 |\n| `references/routes.md` | write a plan, build, check or deliver: what differs by kind |\n| `references/spec.md` | write a vertical video's spec or build it |\n| `references/outline.md` | write a tutorial's outline or build it |\n| `references/script.md` | write a presenter's lines |\n| `references/record.md` | write or resume the project record |\n\n**The challenger at each decision.** Before you go on from a decision, quality-review's challenger (the\n`challenger` agent where the host has it) asks \"why this?\" of it: the set of videos (stage 4), the\nstyle frames (6), the plan and its hooks before the ASK (7), the first build's UI, against the\nrecording's own screens (9), and every quote before its SPEND (8, 12, 13). Act on its verdict as\nquality-review says. A `stop`, or a `rethink` of the user's own choice, goes to the user in the next ASK,\nthe cheaper alternative first, before anything is spent.\n\n## 1. Take the recording in\n\n- **Read the user's brief first.** Every file in the material that is not media, read in full\n  before the first question, as picsart-video-rules's The user's own brief says. A brief sets the message, the\n  audience and the must-shows.\n- **No recording yet → open the upload in your first reply.** When the message carries no recording\n  (no attachment, no link, no Drive file named), call\n  `picsart_drive_upload_ask({ accept: 'video', purpose: 'your screen recording' })` in that same turn,\n  with one line on what happens once it is in, and ask the interview's first question beside it.\n  If the host cannot show the upload, follow picsart-video-rules's host notes.\n- A link the user pasted, or a file in their Picsart Drive: bring it in as picsart-video-rules's Asset URLs\n  say. Every tool below reads the recording by the URL that comes back.\n- **Several recordings** are one set of material: read each, and topics may come from any of them.\n  One video plays one recording.\n\n## 2. The interview\n\nAsk as `references/interview.md` says: one question per turn, in its order, each skipped when the\nuser's words or the recording already answer it. A typical run asks four to six. The questions are\nwhere it goes, what a viewer should get from it, how many videos, who it is for, the message, the\nlength (only when the destination leaves a range), the sound and the ending, and whether the\naccount on screen is real.\n\n- **Start reading at once** (stage 3): the first question goes in the same turn as the read, so the\n  user answers while it runs, and later questions are drafted from what it found (\"your recording\n  shows five things a viewer could learn\").\n- **One question in a turn, never a list.** The only pair is the sound and the ending, as\n  `references/interview.md` marks it. The rest wait for their turn; never preview them. What you\n  settle without asking is said in one line as assumed, so the user can change it.\n- **No plan is written until the message is one sentence.** \"You decide\" counts: write it from the\n  recording and list it as assumed.\n- **The brief closes the interview**: what the user said, what you read, what you assumed, as\n  picsart-video-rules says, with how to change each assumption. The confirmation is the next real decision\n  (stage 4), never a yes.\n\nThe user can say \"just make it\" at any point: take the recommended answers, list them as assumed,\nand go on to the next choice made by eye or the next charge.\n\n## 3. Read the recording\n\nRead it as the screen-recording-analysis skill says: the moments first, then still sheets of the moments that\nmatter, and for a long session its stages. The recording is never sent to a model that watches\nvideo: never call `picsart_media_describe_video`. footage-analysis is used only for the transcript.\n\nThe read is the moments, the still sheets (about four per recording, one sheet of candidate results\nfor the set), the stages where they are needed, and the transcript when `input.hasAudio` and the\nvoice may be kept. It runs no command: `scripts/actions.py` waits for the outline (stage 7). Where\nyou cannot run commands, never look for a shell or another way to run a script; read with the\ntools alone, and the build stops as stage 9 says. Say in one line that you are reading, and go on; a call that charges goes through\npicsart-video-rules's SPEND first.\n\nIf a read fails twice, tell the user the service is down and to try again shortly. Never guess what\nthe recording shows. **Moments already read in this session are not read again.**\n\nWhat the read decides, said in one line each and never asked:\n\n- **The shape.** `input.orientation` sets the layout, as the screen-recording-layouts skill says: a portrait\n  recording fills the frame, a square one fills its width. A tutorial needs a raw landscape\n  recording; a portrait or edited one gets the vertical kinds instead (`references/routes.md`).\n- **The tier**, for a promo: prompt over result, timelapse or re-cut, from `tier` and `tierReason`,\n  checked on the first sheet. When the sheet disagrees, believe the sheet.\n- **What the recording is:** the product, what its steps get done, its own colours (for the\n  style's brand), and whether there is speech.\n\n## 4. The set: which videos\n\nFind the candidates as `references/routes.md` says for the kind: standalone topics for how-to\nshorts (each a run of actions from a state a viewer can reach to a visible result), the features\nfor an explainer, the steps that sell for a promo. **Every candidate comes from a still you read.**\n\nA result that is only words in a chat reply, or that comes after the recording ends, is not on a\nstill: leave that candidate out, and say in one line what footage would bring it back.\n\nSHOW one sheet of the candidates' results, then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which to make, as `references/interview.md`\nsays: a multi-select, the most teachable first, each with what the viewer learns and where it is\nin the recording. A tutorial skips this stage: it keeps every step, and its outline is stage 7.\n\nEvery short stands alone: its own hook, its own payoff, its own ending. Never \"part 2\", never a\nreference to another short.\n\n**Will the screens read?** For the vertical kinds, once the videos are chosen, measure each moment\nthey will show with the screen-recording-layouts skill's `readability.py`: its box, the smallest text the viewer\nmust read and the words in it (from the still sheets), its event's luma, and the recommended\nstyle's canvas. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) \"The screens\" (`references/interview.md`), its own question, with the\nmeasured reason: when a moment fails, or when the kind is a how-to or a promo where the UI is only\nthe means. A tutorial keeps the recording: it teaches where to click in the real app.\n\n## 5. How we iterate\n\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what the user wants to see first (`references/interview.md`): stills of video 1 in three\nstyles, video 1 finished, or the written plan of every video. Then say the agreed plan in one line,\nas picsart-video-rules's Asking says, with what costs: \"Then: you pick a style on three stills, I build\nshort 1, you review it (free, as many rounds as you like), then I build the other four in the same\nstyle and you review them together. Only the voice costs, quoted first.\" Store it in the record\nand repeat its line at each pass.\n\n## 6. The style\n\nChoose it as the visual-style skill says. Recommend one from the recording and the brief, with its\nreason; build video 1's hook frame and one step frame in it and in two styles that differ, with the\nrecording's colours as the brand where they fit; and SHOW them as three contact sheets on the user's\nown recording. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which style, the recommended one first. Building and the contact sheet are\nfree. A mix (\"the colours of one, the type of the other\") comes through Other: build it and show\nit.\n\nThe pick is the set's style. Save it in the record; every video of the set uses it.\n\n## 7. The plan of video 1\n\nWrite the plan as `references/routes.md` says for the kind. It opens with one sentence, \"This\nshort shows <audience> that <message>\", and has one row per beat: the time, what is on screen, the\nwords, and why the beat is there, traced to the message. Write the words as the spec's or the\noutline's words rules say and the retention-hooks skill says. **Never invent what the recording does\nnot show.**\n\nWith the plan, write its viewer track, as quality-review's cold watch says: the promise by 3 s,\neach step a newcomer must follow, and the payoff, with the questions to ask after watching. Its\nquestions are shown with the rows, for correction.\n\n**Review the script before anyone sees the plan**, as quality-review's script review says: put its\nrewritten lines into the rows, and its why under the opening sentence. A `drop` is not planned:\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what to do with that video, with the reason quoted: \"Leave it out\" (recommended: the\nrecording never shows its result) / \"Teach the smaller lesson\" (the one the review names) / \"I'll\nrecord the result\" (the footage its `rescue` names). The review is free, so it never asks to run.\n\nThe ASK is the hook: 2-3 drafted openings that differ in what they open on (a problem the viewer\nfeels, the result in the middle of happening, a before and after), each written as the retention-hooks\nskill's hook rules say, each saying what it opens on. The rows are shown for correction, not for\napproval: anything the user writes changes them, and their answer to the hook moves on.\n\n**Each hook built, before the ASK** (a vertical video). Build each drafted hook, which is free,\ntake its first 2 s on a contact sheet, and check it against the retention-hooks skill's hook rules.\nRewrite one that breaks them. Offer each as a prediction, never as tested: nobody has watched it\nyet. A hook that opens on footage not made yet is judged on its still or its description. The\ndraft's cold watch is the hook's test: its first two pieces (to 3 s) say whether strangers keep\nwatching.\n\nAn opening the recording does not hold, such as a generated shot, is a SPEND. **The first reply\nabout a generated shot offers its key frames and nothing else:** stills at the hook, the turn and\nthe payoff, quoted on their own, shown together before any video is priced, and reused to open the\nshot. No video model is read, compiled or priced until the user approves the stills. **Pictures the\nuser brings, a film's scene pictures included, are references for those stills, never the stills\nthemselves**: they are not this short's frame or its beats, so the key frames are still drafted\nfrom them, quoted and shown first, even when the user says to go ahead. That reply says all four\nto the user, in its own words, beside the quote:\n\n- **Every key point**: a still for each generated beat (a generated hook and payoff are two stills,\n  one batch, one quote), and the recording's own frame at the turn, which is free.\n- **The quote covers those stills alone.** No video price, take or model in it.\n- **They are shown together** as the concept before any video is priced.\n- **Each approved still opens its shot** as its start frame; none is a throwaway preview.\n\nFor example: \"You'll see the hook and payoff stills together, beside the recording's frame at the\nturn, before any video is priced; each one you approve becomes the first frame of its shot.\"\n\nAn approved plan already names its hook, turn and payoff, so this reply runs no script: take the\nkey points from the plan, and never search for a shell or hand the script to another agent. Draft it first and\nquote the draft beside the final, as picsart-video-rules's SPEND says; a crop of the recording is usually\nsmaller than the frame, so say so when it is the reference.\nIts draft is key frames, as picsart-video-rules's SPEND rule 8 says: plan them on this plan's beats with the\nfilm-shot-generation skill's `scripts/keyframes.py`, SHOW them before the shot is quoted, and open the shot on\nthe approved one. In the hook ASK, a generated opening is offered as its key frames first.\nIts key frames and its shot are compiled with `picsart_film_compile_prompt` (picsart-video-rules's SPEND rule 9), with\nthe size and lens the recording's own plan gives that shot.\n\nThe words' language is the user's own, unless the recording's words are in another: then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which.\n\n## 8. The voice, when chosen\n\nThe sound was chosen in the interview. Make it as `references/routes.md` says:\n\n- **A generated voice**: one line per beat, quoted as one batch with `picsart_preflight` after\n  reading the voice model with `picsart_model_params` (picsart-video-rules, SPEND). Generate after the yes,\n  probe each line's real length, and transcribe each for its captions, as the voice-over-generation and\n  spoken-captions skills say. The line sets the beat's length; the picture holds.\n- **The user's own voice** (`input.hasAudio`, a transcript that is not thin): kept whole, never cut\n  mid-sentence, captioned from its transcript as `picsart_media_transcribe` returned it.\n- **A presenter on camera**: the promo route only, never the default, as `references/routes.md`\n  says.\n- **Captions and music, or text only**: nothing to generate here.\n\nSeveral videos' voice lines are one batch: one quote and one yes for the set.\n\n## 9. Build\n\nBuild with the script for the kind, from this use case's folder. Run it; never work the scene out\nyourself.\n\n1. Write its input to a file as `references/spec.md` (vertical kinds) or `references/outline.md`\n   (tutorial) says, with `style` and `brand` from stage 6 in the spec, `clip: \"mask\"`, and\n   `moments` exactly as `picsart_media_screen_moments` returned them.\n2. Run `python3 scripts/promo_scene.py <file>` or `python3 scripts/tutorial_scene.py <file>`. It\n   prints one JSON line with the scene, its length, a timeline and diagnostics. The skills'\n   scripts it imports ship beside it in `scripts/`.\n3. If it exits with an error instead, its one line names the field it cannot build from. Fix that\n   field and run it again.\n\nBuilding is free and calls no tool. **Where you cannot run Python**, the build stops as\npicsart-video-rules says for a script that builds the scene: say this video needs the scene builder,\noffer to carry on on the desktop or in the Picsart app from the record, and stop. Never hand-build it.\n\nA timelapse omits stages with no usable footage inside their source bounds and cut guards.\n\n**Act on every diagnostic before you look at anything else.** An `error` means the scene is wrong:\nmake the change its `fix` names and build again; never patch the scene past one. A `warning` means\nit plays but may read wrong: fix it when the fix keeps what the user chose, otherwise look at that\nmoment at the check and say it in the review line. An `error` still there after two builds: tell\nthe user what it says and what you changed, and stop there.\n\n**A step to recreate** (the user chose designed UI for it): after the build, write the step's\nelements and states from the still notes as the screen-ui-redesign skill says, with the step's panel as\n`box` and its timeline span as `start` and `duration`, and run its `scripts/ui_beat.py`. With\n`picsart_scene_patch`, remove the step's `<id>_panel` and `<id>_rim` and add the beat's\nlayers and assets in their place. Do it again after every build, as the covers are. The redraw is\nthe app's own screen, larger and cleaner, never an invented one: before the video is rendered,\ncompare each state's frame with its `source` still, as the screen-ui-redesign skill says.\n\nCaptions, a presenter's takes and callouts go on after the build with\n`picsart_scene_patch`, as `references/routes.md` says. The build's pushes are its own; a\npush you add by hand, on a scene the user edited, comes from the screen-recording-zoom skill's\n`zoom_plan.py`, `balanced` unless the user asked for gentler or punchier pushes. When the user's comments change what a\nvideo shows or says, change the spec or outline and build again; when they edited the scene in the\neditor, their document replaces yours (picsart-video-rules): patch it instead.\n\n**A generated shot goes in as a presenter's take does**: build with `hook.duration` set to its\nlength, then patch it in full frame with `picsart_scene_patch` (`references/routes.md`).\nNever swap the script for a scene of your own. A scene worked out by hand skips the build's\ndiagnostics, and the stages around it still hold: key frames and SPEND before, the check and REVIEW\nafter.\n\n## 10. Check it\n\nValidate the build with `picsart_scene_validate`, then check it with quality-review, at most 8\nmoments to a contact sheet: the hook, the end of each beat where its push has landed, the ending,\nand the moment of each warning left in. Ask the questions `references/routes.md` gives for the\nkind. Then the polish pass, which quality-review runs:\n\n- the build's colours and entrances keep to its style (visual-style's `adherence.py`);\n- nothing on the visual-style skill's list of defaults that look machine-made, unless it is\n  kept with a reason;\n- every word and the call to action sit inside the destination's safe zone: the screen-recording-layouts skill's\n  `safe_zone.py`, with the profile of each place in the brief;\n- every line is held long enough to read and the picture never stalls (quality-review's `timing.py`,\n  with each step's measured reveal as a change);\n- the first frame is a real moment (the payoff, or the problem it solves), not a login screen or\n  an empty app;\n- a vertical video's draft gets quality-review's cold watch against its viewer track, quoted first\n  as it says, with its SPEND (quote it in one line, then get the yes as an ASK with \"Yes, go ahead\" / \"Yes, up to N credits for this task\" / \"No\": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit (\"as much as needed\"), never inferred from words like \"go ahead\". Within it, ask nothing: quote it in one line (\"About 8 credits; 12 of 20 left after this\"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) before round 1 unless every call is quoted at 0 credits: its first\n  two pieces test the hook, and its floor line goes in the review line. When the user says no, the\n  review line says \"no cold watch yet\";\n- then quality-review's critic panel runs on every draft (the six `critic-*` agents where the host has\n  them), with the recording's stills at each step as its source and the cold watch's gap list. Make its must-fixes, and ASK its `choices` before REVIEW, never applied silently.\n\nThen grade the brief with quality-review's `checklist.py`: the message, the destinations, the length,\nthe hook and the ending from the record, and the plan's step lines. Fix only what it says no to.\n\n**Private data, before anything is shown.** Run the screen-recording-analysis skill's privacy pass on every\ntext you read for this video, on its stills and in its transcript. With real data, cover each\nhit it keeps, and each name or address you saw, with the screen-recording-layouts skill's `cover.py`, in the\nstyle's surface colour, and patch the cards in; with a demo account, the keys, tokens and card\nnumbers only. A box on screen for a whole step can be cut or framed out instead. Cover again after\nevery build: a build drops the patched cards. Never open REVIEW with a hit uncovered; say in the\nreview line what was covered.\n\n**Masks first.** The first contact sheet must include a panel. If it fails with `the authored mask\ncarries no clip`, masks are not rendering here (the screen-recording-layouts skill has the dated fact): build the\nsame spec again with `clip: \"nest\"` and check that build. REVIEW opens the mask build; the nest\nbuild is the one that renders.\n\n**Judge legibility on a still, not on softness alone**: crop a `picsart_media_video_stills` sheet\nto the beat's box at that moment. If the still is sharp, the recording is legible; never push\nharder to fix a soft contact sheet (screen-recording-zoom).\n\n## 11. The user reviews\n\nREVIEW the mask build, titled \"<Product> <kind> — draft 1\", opened cold as picsart-video-rules\nsays: its length and the track's questions. Once the user has answered, give the checklist's\n`summary`, the cold watch's floor line and the line `references/routes.md` gives for the kind (\"Does\nthe result land in the first two seconds? Is each click readable? Say where it drags\"). Each new\ndraft's number goes up by one. Comments apply\nwhere they point; after the third round on the same video, ask what is causing the misses (the\nlook, the pace or the words) instead of patching again.\n\n## 12. The rest of the set\n\nBuild the other videos with video 1's approved style and with what the user's comments on it\ntaught (the pace, the words' tone, the moves). Plan and review each as stage 7 does, with every `drop` in the set asked in one turn, and its own hook asked\nonly when the user chose to see plans; otherwise draft the hook yourself and list it as assumed.\nTheir voice lines are one batch SPEND. REVIEW them one after another, and ask one \"anything to\nchange?\" for the set (picsart-video-rules).\n\n## 13. Music, when wanted\n\nOffer it once, after the drafts are approved, unless the sound chosen was the user's own voice or\nthe recording's own sound. ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) with two drafted briefs that differ in feel, as the background-music skill\nsays, and \"No music\". On a pick, the bed (and a promo's push sounds, `references/routes.md`) is one\nSPEND (quote it in one line, then get the yes as an ASK with \"Yes, go ahead\" / \"Yes, up to N credits for this task\" / \"No\": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit (\"as much as needed\"), never inferred from words like \"go ahead\". Within it, ask nothing: quote it in one line (\"About 8 credits; 12 of 20 left after this\"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) for the set. Lay it under each video as the background-music skill says, validate and look as stage\n10 does, and REVIEW, titled \"<Product> <kind> — music\", with one line on what to listen for. If it\nfails or is declined, say so in one line and go on with the approved drafts.\n\n## 14. Deliver\n\nDELIVER each approved video: the mask build if masks rendered at the check, otherwise the nest\nbuild. Check each export against its scene with quality-review's `qc.py` before you give its link. If the user edited a scene in the editor and masks do not render, export their scene; if\nthat fails on its masks, say so and offer the nest build, naming the edits it leaves out.\n\nWith each video, give:\n\n- the link, its length against the target, and its status: done, done with concerns (each named),\n  or blocked;\n- **post copy** for the destination, as `references/routes.md` says: a title under 40 characters,\n  a first caption line that hooks and a caption that says what the video shows, and 3-5 words of\n  cover text taken from the payoff frame; hashtags only if asked;\n- what it sounds like: silent (a track can be added when it is posted), the user's voice, a\n  generated voice (AI-made: tell them to switch on the platform's AI-content label), or music;\n- a tutorial's chapters, exactly as the build printed them, to paste into the description.\n\nThen offer one next step in a sentence: another kind from the same recording (nothing is read\nagain), or saving this set's brief and style in the record so \"make another\" starts from them.\n"
Full snapshot data
{
  "description": "Turn a screen recording of an app or site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with chapters - in a style picked by eye, with a voice-over and captions if wanted. Use when the user has a screen recording and wants a tutorial video (for a help center too), a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including a generated shot. Not for camera footage or a text-only article.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 185
    },
    {
      "relative_path": "references/interview.md",
      "size_in_bytes": 9893
    },
    {
      "relative_path": "references/outline.md",
      "size_in_bytes": 9217
    },
    {
      "relative_path": "references/record.md",
      "size_in_bytes": 1849
    },
    {
      "relative_path": "references/routes.md",
      "size_in_bytes": 12444
    },
    {
      "relative_path": "references/script.md",
      "size_in_bytes": 2023
    },
    {
      "relative_path": "references/spec.md",
      "size_in_bytes": 15214
    },
    {
      "relative_path": "scripts/_build.py",
      "size_in_bytes": 33482
    },
    {
      "relative_path": "scripts/_diagnostics.py",
      "size_in_bytes": 21558
    },
    {
      "relative_path": "scripts/_geometry.py",
      "size_in_bytes": 1139
    },
    {
      "relative_path": "scripts/_input.py",
      "size_in_bytes": 12663
    },
    {
      "relative_path": "scripts/_js.py",
      "size_in_bytes": 4625
    },
    {
      "relative_path": "scripts/_moments.py",
      "size_in_bytes": 1497
    },
    {
      "relative_path": "scripts/_preview.py",
      "size_in_bytes": 5743
    },
    {
      "relative_path": "scripts/_scene.py",
      "size_in_bytes": 21419
    },
    {
      "relative_path": "scripts/_stages.py",
      "size_in_bytes": 5261
    },
    {
      "relative_path": "scripts/_steps.py",
      "size_in_bytes": 1687
    },
    {
      "relative_path": "scripts/_timeedit.py",
      "size_in_bytes": 43226
    },
    {
      "relative_path": "scripts/_tutorial_edit.py",
      "size_in_bytes": 14003
    },
    {
      "relative_path": "scripts/_tutorial_input.py",
      "size_in_bytes": 9273
    },
    {
      "relative_path": "scripts/_tutorial_layout.py",
      "size_in_bytes": 12483
    },
    {
      "relative_path": "scripts/actions.py",
      "size_in_bytes": 11047
    },
    {
      "relative_path": "scripts/adherence.py",
      "size_in_bytes": 10597
    },
    {
      "relative_path": "scripts/cover.py",
      "size_in_bytes": 16819
    },
    {
      "relative_path": "scripts/framing.py",
      "size_in_bytes": 7644
    },
    {
      "relative_path": "scripts/panels.py",
      "size_in_bytes": 13591
    },
    {
      "relative_path": "scripts/private_text.py",
      "size_in_bytes": 17020
    },
    {
      "relative_path": "scripts/promo_scene.py",
      "size_in_bytes": 2889
    },
    {
      "relative_path": "scripts/readability.py",
      "size_in_bytes": 7792
    },
    {
      "relative_path": "scripts/safe_zone.py",
      "size_in_bytes": 61049
    },
    {
      "relative_path": "scripts/style.py",
      "size_in_bytes": 7576
    },
    {
      "relative_path": "scripts/tutorial_scene.py",
      "size_in_bytes": 17091
    },
    {
      "relative_path": "scripts/zoom.py",
      "size_in_bytes": 15530
    },
    {
      "relative_path": "scripts/zoom_plan.py",
      "size_in_bytes": 16698
    }
  ],
  "name": "make-screen-videos",
  "skill_md_contents": "---\nname: make-screen-videos\ndescription: Turn a screen recording of an app or site into videos - standalone how-to shorts for Reels, TikTok or Shorts each teaching one easy thing, a feature explainer, a promo with each request typed over its result, a long session's timelapse, or a step-by-step 16:9 tutorial with chapters - in a style picked by eye, with a voice-over and captions if wanted. Use when the user has a screen recording and wants a tutorial video (for a help center too), a how-to, a reel, a short, a promo, a walkthrough or a demo video from it, or carries on with videos already planned from one, including a generated shot. Not for camera footage or a text-only article.\nmetadata:\n  uses: screen-recording-analysis, footage-analysis, speech-aware-cutting, screen-recording-layouts, screen-recording-zoom, visual-style, screen-ui-redesign, retention-hooks, scene-building-basics, quality-review, ai-presenter, voice-over-generation, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  optional: ai-presenter, voice-over-generation, speech-aware-cutting, spoken-captions, background-music, motion-design-craft, film-shot-generation\n  imports: screen-recording-analysis, screen-recording-layouts, screen-recording-zoom, visual-style\n  title: Make screen videos\n  short-description: Turn a screen recording into how-to videos\n  argument-hint: '[screen recording and the video you want]'\n  starter-prompt: Turn my screen recording into short how-to videos.\n---\n\n# Screen video\n\nA screen recording is mostly waiting, and what it shows is a few pixels of dense UI. The same\nrecording can become very different videos: a set of 20-second shorts that each show how easy one\nthing is, a promo that sells, or a tutorial that keeps every step. Which one is the user's to say,\nso this use case starts with a short interview, never with a draft to approve.\n\nThe kinds, each a route in `references/routes.md`:\n\n| Kind | For | Frame | Built by |\n|---|---|---|---|\n| **How-to short** | one easy task, start to result: \"look how easy\" | 9:16 | `scripts/promo_scene.py` |\n| **Feature explainer** | one feature: the problem, the feature, two or three uses | 9:16 or 16:9 | `scripts/promo_scene.py` |\n| **Promo** | the best results, fast, to sell; a timelapse for a long session; a re-cut for an edited video | 9:16 | `scripts/promo_scene.py` |\n| **Tutorial** | every step in order, held long enough to follow, with chapters | 16:9 | `scripts/tutorial_scene.py` |\n\nRead picsart-video-rules first: how to talk and how to ask are there, and every question here follows its\nAsking rules. One job: one or more screen recordings in; one video, or a set of standalone videos,\nout.\n\nBefore acting on a stage, load the skills in its row that apply to this job, also when the work resumes there.\nA gate with a scope in brackets opens only there: `SPEND (vertical)` is the cold watch's go-ahead,\nfor a vertical video's draft alone; a horizontal video goes straight on through stage 10.\n\n| Stage | Gate | Skills |\n|---|---|---|\n| 1. Take the recording in | — | — |\n| 2. The interview | ASK | — |\n| 3. Read the recording | — | screen-recording-analysis, footage-analysis |\n| 4. The set: which videos | SHOW, ASK | screen-recording-analysis, screen-recording-layouts |\n| 5. How we iterate | ASK | — |\n| 6. The style | SHOW, ASK | visual-style, screen-recording-layouts |\n| 7. The plan of video 1 | ASK | retention-hooks, quality-review, film-shot-generation |\n| 8. The voice, when chosen | ASK, SPEND | voice-over-generation, speech-aware-cutting, spoken-captions, ai-presenter |\n| 9. Build | — | screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style, screen-ui-redesign, motion-design-craft |\n| 10. Check it | SPEND (vertical) | quality-review, visual-style, screen-recording-zoom, screen-recording-layouts |\n| 11. The user reviews | REVIEW | — |\n| 12. The rest of the set | SPEND, REVIEW | retention-hooks, quality-review, voice-over-generation, screen-recording-layouts, screen-recording-zoom, scene-building-basics, visual-style |\n| 13. Music, when wanted | ASK, SPEND, REVIEW | background-music |\n| 14. Deliver | DELIVER | — |\n\n| Read this | Before you |\n|---|---|\n| `references/interview.md` | ask the first question, and at every ASK in stages 2-7 |\n| `references/routes.md` | write a plan, build, check or deliver: what differs by kind |\n| `references/spec.md` | write a vertical video's spec or build it |\n| `references/outline.md` | write a tutorial's outline or build it |\n| `references/script.md` | write a presenter's lines |\n| `references/record.md` | write or resume the project record |\n\n**The challenger at each decision.** Before you go on from a decision, quality-review's challenger (the\n`challenger` agent where the host has it) asks \"why this?\" of it: the set of videos (stage 4), the\nstyle frames (6), the plan and its hooks before the ASK (7), the first build's UI, against the\nrecording's own screens (9), and every quote before its SPEND (8, 12, 13). Act on its verdict as\nquality-review says. A `stop`, or a `rethink` of the user's own choice, goes to the user in the next ASK,\nthe cheaper alternative first, before anything is spent.\n\n## 1. Take the recording in\n\n- **Read the user's brief first.** Every file in the material that is not media, read in full\n  before the first question, as picsart-video-rules's The user's own brief says. A brief sets the message, the\n  audience and the must-shows.\n- **No recording yet → open the upload in your first reply.** When the message carries no recording\n  (no attachment, no link, no Drive file named), call\n  `picsart_drive_upload_ask({ accept: 'video', purpose: 'your screen recording' })` in that same turn,\n  with one line on what happens once it is in, and ask the interview's first question beside it.\n  If the host cannot show the upload, follow picsart-video-rules's host notes.\n- A link the user pasted, or a file in their Picsart Drive: bring it in as picsart-video-rules's Asset URLs\n  say. Every tool below reads the recording by the URL that comes back.\n- **Several recordings** are one set of material: read each, and topics may come from any of them.\n  One video plays one recording.\n\n## 2. The interview\n\nAsk as `references/interview.md` says: one question per turn, in its order, each skipped when the\nuser's words or the recording already answer it. A typical run asks four to six. The questions are\nwhere it goes, what a viewer should get from it, how many videos, who it is for, the message, the\nlength (only when the destination leaves a range), the sound and the ending, and whether the\naccount on screen is real.\n\n- **Start reading at once** (stage 3): the first question goes in the same turn as the read, so the\n  user answers while it runs, and later questions are drafted from what it found (\"your recording\n  shows five things a viewer could learn\").\n- **One question in a turn, never a list.** The only pair is the sound and the ending, as\n  `references/interview.md` marks it. The rest wait for their turn; never preview them. What you\n  settle without asking is said in one line as assumed, so the user can change it.\n- **No plan is written until the message is one sentence.** \"You decide\" counts: write it from the\n  recording and list it as assumed.\n- **The brief closes the interview**: what the user said, what you read, what you assumed, as\n  picsart-video-rules says, with how to change each assumption. The confirmation is the next real decision\n  (stage 4), never a yes.\n\nThe user can say \"just make it\" at any point: take the recommended answers, list them as assumed,\nand go on to the next choice made by eye or the next charge.\n\n## 3. Read the recording\n\nRead it as the screen-recording-analysis skill says: the moments first, then still sheets of the moments that\nmatter, and for a long session its stages. The recording is never sent to a model that watches\nvideo: never call `picsart_media_describe_video`. footage-analysis is used only for the transcript.\n\nThe read is the moments, the still sheets (about four per recording, one sheet of candidate results\nfor the set), the stages where they are needed, and the transcript when `input.hasAudio` and the\nvoice may be kept. It runs no command: `scripts/actions.py` waits for the outline (stage 7). Where\nyou cannot run commands, never look for a shell or another way to run a script; read with the\ntools alone, and the build stops as stage 9 says. Say in one line that you are reading, and go on; a call that charges goes through\npicsart-video-rules's SPEND first.\n\nIf a read fails twice, tell the user the service is down and to try again shortly. Never guess what\nthe recording shows. **Moments already read in this session are not read again.**\n\nWhat the read decides, said in one line each and never asked:\n\n- **The shape.** `input.orientation` sets the layout, as the screen-recording-layouts skill says: a portrait\n  recording fills the frame, a square one fills its width. A tutorial needs a raw landscape\n  recording; a portrait or edited one gets the vertical kinds instead (`references/routes.md`).\n- **The tier**, for a promo: prompt over result, timelapse or re-cut, from `tier` and `tierReason`,\n  checked on the first sheet. When the sheet disagrees, believe the sheet.\n- **What the recording is:** the product, what its steps get done, its own colours (for the\n  style's brand), and whether there is speech.\n\n## 4. The set: which videos\n\nFind the candidates as `references/routes.md` says for the kind: standalone topics for how-to\nshorts (each a run of actions from a state a viewer can reach to a visible result), the features\nfor an explainer, the steps that sell for a promo. **Every candidate comes from a still you read.**\n\nA result that is only words in a chat reply, or that comes after the recording ends, is not on a\nstill: leave that candidate out, and say in one line what footage would bring it back.\n\nSHOW one sheet of the candidates' results, then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which to make, as `references/interview.md`\nsays: a multi-select, the most teachable first, each with what the viewer learns and where it is\nin the recording. A tutorial skips this stage: it keeps every step, and its outline is stage 7.\n\nEvery short stands alone: its own hook, its own payoff, its own ending. Never \"part 2\", never a\nreference to another short.\n\n**Will the screens read?** For the vertical kinds, once the videos are chosen, measure each moment\nthey will show with the screen-recording-layouts skill's `readability.py`: its box, the smallest text the viewer\nmust read and the words in it (from the still sheets), its event's luma, and the recommended\nstyle's canvas. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) \"The screens\" (`references/interview.md`), its own question, with the\nmeasured reason: when a moment fails, or when the kind is a how-to or a promo where the UI is only\nthe means. A tutorial keeps the recording: it teaches where to click in the real app.\n\n## 5. How we iterate\n\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what the user wants to see first (`references/interview.md`): stills of video 1 in three\nstyles, video 1 finished, or the written plan of every video. Then say the agreed plan in one line,\nas picsart-video-rules's Asking says, with what costs: \"Then: you pick a style on three stills, I build\nshort 1, you review it (free, as many rounds as you like), then I build the other four in the same\nstyle and you review them together. Only the voice costs, quoted first.\" Store it in the record\nand repeat its line at each pass.\n\n## 6. The style\n\nChoose it as the visual-style skill says. Recommend one from the recording and the brief, with its\nreason; build video 1's hook frame and one step frame in it and in two styles that differ, with the\nrecording's colours as the brand where they fit; and SHOW them as three contact sheets on the user's\nown recording. Then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which style, the recommended one first. Building and the contact sheet are\nfree. A mix (\"the colours of one, the type of the other\") comes through Other: build it and show\nit.\n\nThe pick is the set's style. Save it in the record; every video of the set uses it.\n\n## 7. The plan of video 1\n\nWrite the plan as `references/routes.md` says for the kind. It opens with one sentence, \"This\nshort shows <audience> that <message>\", and has one row per beat: the time, what is on screen, the\nwords, and why the beat is there, traced to the message. Write the words as the spec's or the\noutline's words rules say and the retention-hooks skill says. **Never invent what the recording does\nnot show.**\n\nWith the plan, write its viewer track, as quality-review's cold watch says: the promise by 3 s,\neach step a newcomer must follow, and the payoff, with the questions to ask after watching. Its\nquestions are shown with the rows, for correction.\n\n**Review the script before anyone sees the plan**, as quality-review's script review says: put its\nrewritten lines into the rows, and its why under the opening sentence. A `drop` is not planned:\nASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) what to do with that video, with the reason quoted: \"Leave it out\" (recommended: the\nrecording never shows its result) / \"Teach the smaller lesson\" (the one the review names) / \"I'll\nrecord the result\" (the footage its `rescue` names). The review is free, so it never asks to run.\n\nThe ASK is the hook: 2-3 drafted openings that differ in what they open on (a problem the viewer\nfeels, the result in the middle of happening, a before and after), each written as the retention-hooks\nskill's hook rules say, each saying what it opens on. The rows are shown for correction, not for\napproval: anything the user writes changes them, and their answer to the hook moves on.\n\n**Each hook built, before the ASK** (a vertical video). Build each drafted hook, which is free,\ntake its first 2 s on a contact sheet, and check it against the retention-hooks skill's hook rules.\nRewrite one that breaks them. Offer each as a prediction, never as tested: nobody has watched it\nyet. A hook that opens on footage not made yet is judged on its still or its description. The\ndraft's cold watch is the hook's test: its first two pieces (to 3 s) say whether strangers keep\nwatching.\n\nAn opening the recording does not hold, such as a generated shot, is a SPEND. **The first reply\nabout a generated shot offers its key frames and nothing else:** stills at the hook, the turn and\nthe payoff, quoted on their own, shown together before any video is priced, and reused to open the\nshot. No video model is read, compiled or priced until the user approves the stills. **Pictures the\nuser brings, a film's scene pictures included, are references for those stills, never the stills\nthemselves**: they are not this short's frame or its beats, so the key frames are still drafted\nfrom them, quoted and shown first, even when the user says to go ahead. That reply says all four\nto the user, in its own words, beside the quote:\n\n- **Every key point**: a still for each generated beat (a generated hook and payoff are two stills,\n  one batch, one quote), and the recording's own frame at the turn, which is free.\n- **The quote covers those stills alone.** No video price, take or model in it.\n- **They are shown together** as the concept before any video is priced.\n- **Each approved still opens its shot** as its start frame; none is a throwaway preview.\n\nFor example: \"You'll see the hook and payoff stills together, beside the recording's frame at the\nturn, before any video is priced; each one you approve becomes the first frame of its shot.\"\n\nAn approved plan already names its hook, turn and payoff, so this reply runs no script: take the\nkey points from the plan, and never search for a shell or hand the script to another agent. Draft it first and\nquote the draft beside the final, as picsart-video-rules's SPEND says; a crop of the recording is usually\nsmaller than the frame, so say so when it is the reference.\nIts draft is key frames, as picsart-video-rules's SPEND rule 8 says: plan them on this plan's beats with the\nfilm-shot-generation skill's `scripts/keyframes.py`, SHOW them before the shot is quoted, and open the shot on\nthe approved one. In the hook ASK, a generated opening is offered as its key frames first.\nIts key frames and its shot are compiled with `picsart_film_compile_prompt` (picsart-video-rules's SPEND rule 9), with\nthe size and lens the recording's own plan gives that shot.\n\nThe words' language is the user's own, unless the recording's words are in another: then ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) which.\n\n## 8. The voice, when chosen\n\nThe sound was chosen in the interview. Make it as `references/routes.md` says:\n\n- **A generated voice**: one line per beat, quoted as one batch with `picsart_preflight` after\n  reading the voice model with `picsart_model_params` (picsart-video-rules, SPEND). Generate after the yes,\n  probe each line's real length, and transcribe each for its captions, as the voice-over-generation and\n  spoken-captions skills say. The line sets the beat's length; the picture holds.\n- **The user's own voice** (`input.hasAudio`, a transcript that is not thin): kept whole, never cut\n  mid-sentence, captioned from its transcript as `picsart_media_transcribe` returned it.\n- **A presenter on camera**: the promo route only, never the default, as `references/routes.md`\n  says.\n- **Captions and music, or text only**: nothing to generate here.\n\nSeveral videos' voice lines are one batch: one quote and one yes for the set.\n\n## 9. Build\n\nBuild with the script for the kind, from this use case's folder. Run it; never work the scene out\nyourself.\n\n1. Write its input to a file as `references/spec.md` (vertical kinds) or `references/outline.md`\n   (tutorial) says, with `style` and `brand` from stage 6 in the spec, `clip: \"mask\"`, and\n   `moments` exactly as `picsart_media_screen_moments` returned them.\n2. Run `python3 scripts/promo_scene.py <file>` or `python3 scripts/tutorial_scene.py <file>`. It\n   prints one JSON line with the scene, its length, a timeline and diagnostics. The skills'\n   scripts it imports ship beside it in `scripts/`.\n3. If it exits with an error instead, its one line names the field it cannot build from. Fix that\n   field and run it again.\n\nBuilding is free and calls no tool. **Where you cannot run Python**, the build stops as\npicsart-video-rules says for a script that builds the scene: say this video needs the scene builder,\noffer to carry on on the desktop or in the Picsart app from the record, and stop. Never hand-build it.\n\nA timelapse omits stages with no usable footage inside their source bounds and cut guards.\n\n**Act on every diagnostic before you look at anything else.** An `error` means the scene is wrong:\nmake the change its `fix` names and build again; never patch the scene past one. A `warning` means\nit plays but may read wrong: fix it when the fix keeps what the user chose, otherwise look at that\nmoment at the check and say it in the review line. An `error` still there after two builds: tell\nthe user what it says and what you changed, and stop there.\n\n**A step to recreate** (the user chose designed UI for it): after the build, write the step's\nelements and states from the still notes as the screen-ui-redesign skill says, with the step's panel as\n`box` and its timeline span as `start` and `duration`, and run its `scripts/ui_beat.py`. With\n`picsart_scene_patch`, remove the step's `<id>_panel` and `<id>_rim` and add the beat's\nlayers and assets in their place. Do it again after every build, as the covers are. The redraw is\nthe app's own screen, larger and cleaner, never an invented one: before the video is rendered,\ncompare each state's frame with its `source` still, as the screen-ui-redesign skill says.\n\nCaptions, a presenter's takes and callouts go on after the build with\n`picsart_scene_patch`, as `references/routes.md` says. The build's pushes are its own; a\npush you add by hand, on a scene the user edited, comes from the screen-recording-zoom skill's\n`zoom_plan.py`, `balanced` unless the user asked for gentler or punchier pushes. When the user's comments change what a\nvideo shows or says, change the spec or outline and build again; when they edited the scene in the\neditor, their document replaces yours (picsart-video-rules): patch it instead.\n\n**A generated shot goes in as a presenter's take does**: build with `hook.duration` set to its\nlength, then patch it in full frame with `picsart_scene_patch` (`references/routes.md`).\nNever swap the script for a scene of your own. A scene worked out by hand skips the build's\ndiagnostics, and the stages around it still hold: key frames and SPEND before, the check and REVIEW\nafter.\n\n## 10. Check it\n\nValidate the build with `picsart_scene_validate`, then check it with quality-review, at most 8\nmoments to a contact sheet: the hook, the end of each beat where its push has landed, the ending,\nand the moment of each warning left in. Ask the questions `references/routes.md` gives for the\nkind. Then the polish pass, which quality-review runs:\n\n- the build's colours and entrances keep to its style (visual-style's `adherence.py`);\n- nothing on the visual-style skill's list of defaults that look machine-made, unless it is\n  kept with a reason;\n- every word and the call to action sit inside the destination's safe zone: the screen-recording-layouts skill's\n  `safe_zone.py`, with the profile of each place in the brief;\n- every line is held long enough to read and the picture never stalls (quality-review's `timing.py`,\n  with each step's measured reveal as a change);\n- the first frame is a real moment (the payoff, or the problem it solves), not a login screen or\n  an empty app;\n- a vertical video's draft gets quality-review's cold watch against its viewer track, quoted first\n  as it says, with its SPEND (quote it in one line, then get the yes as an ASK with \"Yes, go ahead\" / \"Yes, up to N credits for this task\" / \"No\": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit (\"as much as needed\"), never inferred from words like \"go ahead\". Within it, ask nothing: quote it in one line (\"About 8 credits; 12 of 20 left after this\"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) before round 1 unless every call is quoted at 0 credits: its first\n  two pieces test the hook, and its floor line goes in the review line. When the user says no, the\n  review line says \"no cold watch yet\";\n- then quality-review's critic panel runs on every draft (the six `critic-*` agents where the host has\n  them), with the recording's stills at each step as its source and the cold watch's gap list. Make its must-fixes, and ASK its `choices` before REVIEW, never applied silently.\n\nThen grade the brief with quality-review's `checklist.py`: the message, the destinations, the length,\nthe hook and the ending from the record, and the plan's step lines. Fix only what it says no to.\n\n**Private data, before anything is shown.** Run the screen-recording-analysis skill's privacy pass on every\ntext you read for this video, on its stills and in its transcript. With real data, cover each\nhit it keeps, and each name or address you saw, with the screen-recording-layouts skill's `cover.py`, in the\nstyle's surface colour, and patch the cards in; with a demo account, the keys, tokens and card\nnumbers only. A box on screen for a whole step can be cut or framed out instead. Cover again after\nevery build: a build drops the patched cards. Never open REVIEW with a hit uncovered; say in the\nreview line what was covered.\n\n**Masks first.** The first contact sheet must include a panel. If it fails with `the authored mask\ncarries no clip`, masks are not rendering here (the screen-recording-layouts skill has the dated fact): build the\nsame spec again with `clip: \"nest\"` and check that build. REVIEW opens the mask build; the nest\nbuild is the one that renders.\n\n**Judge legibility on a still, not on softness alone**: crop a `picsart_media_video_stills` sheet\nto the beat's box at that moment. If the still is sharp, the recording is legible; never push\nharder to fix a soft contact sheet (screen-recording-zoom).\n\n## 11. The user reviews\n\nREVIEW the mask build, titled \"<Product> <kind> — draft 1\", opened cold as picsart-video-rules\nsays: its length and the track's questions. Once the user has answered, give the checklist's\n`summary`, the cold watch's floor line and the line `references/routes.md` gives for the kind (\"Does\nthe result land in the first two seconds? Is each click readable? Say where it drags\"). Each new\ndraft's number goes up by one. Comments apply\nwhere they point; after the third round on the same video, ask what is causing the misses (the\nlook, the pace or the words) instead of patching again.\n\n## 12. The rest of the set\n\nBuild the other videos with video 1's approved style and with what the user's comments on it\ntaught (the pace, the words' tone, the moves). Plan and review each as stage 7 does, with every `drop` in the set asked in one turn, and its own hook asked\nonly when the user chose to see plans; otherwise draft the hook yourself and list it as assumed.\nTheir voice lines are one batch SPEND. REVIEW them one after another, and ask one \"anything to\nchange?\" for the set (picsart-video-rules).\n\n## 13. Music, when wanted\n\nOffer it once, after the drafts are approved, unless the sound chosen was the user's own voice or\nthe recording's own sound. ASK (with the host's own choice tool where it offers one, such as request_user_input in Codex's Plan mode, so the user taps an option; a Picsart board the step names asks there instead; only where neither works, the options numbered in one message, the recommended one first; in unattended pace, never ask: write it to the project record's openDecisions with the answer you assumed, as picsart-video-rules says) with two drafted briefs that differ in feel, as the background-music skill\nsays, and \"No music\". On a pick, the bed (and a promo's push sounds, `references/routes.md`) is one\nSPEND (quote it in one line, then get the yes as an ASK with \"Yes, go ahead\" / \"Yes, up to N credits for this task\" / \"No\": the host's own choice tool where it offers one, otherwise those options numbered in one message; the host's own confirmation before a tool runs is not this yes; unless a spend limit the user stated covers it: N credits or an explicit no-limit (\"as much as needed\"), never inferred from words like \"go ahead\". Within it, ask nothing: quote it in one line (\"About 8 credits; 12 of 20 left after this\"), call it, then say what it spent and what remains, keeping a running total that includes failed calls, and stop to ask before the first call that would cross it. A money amount is asked once to restate it in credits; in unattended pace, never ask: call only with a consent or a stated spend limit already recorded, otherwise write the quote to pendingSpends and stop, as picsart-video-rules says) for the set. Lay it under each video as the background-music skill says, validate and look as stage\n10 does, and REVIEW, titled \"<Product> <kind> — music\", with one line on what to listen for. If it\nfails or is declined, say so in one line and go on with the approved drafts.\n\n## 14. Deliver\n\nDELIVER each approved video: the mask build if masks rendered at the check, otherwise the nest\nbuild. Check each export against its scene with quality-review's `qc.py` before you give its link. If the user edited a scene in the editor and masks do not render, export their scene; if\nthat fails on its masks, say so and offer the nest build, naming the edits it leaves out.\n\nWith each video, give:\n\n- the link, its length against the target, and its status: done, done with concerns (each named),\n  or blocked;\n- **post copy** for the destination, as `references/routes.md` says: a title under 40 characters,\n  a first caption line that hooks and a caption that says what the video shows, and 3-5 words of\n  cover text taken from the payoff frame; hashtags only if asked;\n- what it sounds like: silent (a track can be added when it is posted), the user's voice, a\n  generated voice (AI-made: tell them to switch on the platform's AI-content label), or music;\n- a tutorial's chapters, exactly as the build printed them, to paste into the description.\n\nThen offer one next step in a sentence: another kind from the same recording (nothing is read\nagain), or saving this set's brief and style in the record so \"make another\" starts from them.\n"
}

SHA-256 of public snapshot: af8736463bf6dce64ff72d69b524527811f2680b100c89922205d5e8ab54563a