← 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 quality-review

Full description — Now says: draft watched cold by model viewers against its viewer track, then reviewed by

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

short's script, hook and draft reviewed by a simulated viewer panel and

After

Nov 7, 2026 · 00:09 UTC

draft watched cold by model viewers against its viewer track, then reviewed by

Skill instructions

Before

Nov 3, 2026 · 18:14 UTC

short's script, hook and draft reviewed by a simulated viewer panel and critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene,...

After

Nov 7, 2026 · 00:09 UTC

draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a...

Supporting files

Before

Nov 3, 2026 · 18:14 UTC

[{"relative_path":"references/agent-challenger.md","size_in_bytes":3399},{"relative_path":"references/agent-critic-art.md","size_in_bytes":3507},{"relative_path":"references/agent-critic-growth.md","size_in_bytes":3688},{"relative_path":...

After

Nov 7, 2026 · 00:09 UTC

[{"relative_path":"references/agent-challenger.md","size_in_bytes":3399},{"relative_path":"references/agent-critic-art.md","size_in_bytes":3825},{"relative_path":"references/agent-critic-growth.md","size_in_bytes":4201},{"relative_path":...

Compare saved observations

Download comparison JSON

SKILL.md text changes

− removed · + added

--- before
+++ after
@@ -1,6 +1,6 @@
 ---
 name: quality-review
-description: Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a short's script, hook and draft reviewed by a simulated viewer panel and critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.
+description: Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.
 metadata:
   imports: screen-recording-layouts
 ---
@@ -12,14 +12,12 @@
    warnings too.
 2. **Lint the layout.** `picsart_scene_lint_layout` over the timeline; on a long scene, narrow
    `t0`/`t1` to the part in question. Each of its `findings` is a sustained overlap between layers
-   that should not touch. For a pair placed together on purpose, pass `allow: ["idA:idB"]`. Leave
-   `engine` at its default (`native`), which measures real text boxes. A clean lint is not proof:
-   on the `jet` engine it never measures text (a fallback), and on `v3` a text box is the font's
-   line box, so it can flag empty space. If the tool is not in your list, or fails because the
+   that should not touch. For a pair placed together on purpose, pass `allow: ["idA:idB"]`. A text
+   box is the font's line box, so a clean lint is not proof and it can flag empty space. If the tool is not in your list, or fails because the
    query build is unavailable, skip to the frames.
 3. **Measure.** For a flagged pair, or a use-case question about where something sits (inside a
    panel, inside a mask), `picsart_scene_query_layout` at that moment gives each layer's box. It
-   reads layer geometry, not painted pixels, and has no box for text on `jet`.
+   reads layer geometry, not painted pixels.
 4. `picsart_scene_contact_sheet` at each moment the use case names. Look at every frame and answer
    the use case's questions about it.
 5. `picsart_scene_contact_sheet` once across the whole timeline — order, seams, nothing blank.
@@ -63,31 +61,50 @@
 evidence stays open. Fix only the items that say no, then grade again: never rebuild what passed.
 Its `summary` goes with the review.
 
-**The viewer panel.** On a short, before the user sees its hook or its draft, three simulated
-viewers (a 15-year-old scrolling, a creator new to the topic, someone who already uses the
-product) look at the first 2 s and then at the whole draft. Their answers are graded against a
-pass bar with `scripts/panel.py`. It catches a hook that reads as an ad or promises the wrong
-thing. It is not a measure of retention, and it says so.
+**The cold watch.** Before the critics, and before the user sees a draft that has a viewer track,
+three model viewers who know nothing of it watch it forward, a few seconds at a time, from its
+frames and its heard words, each keeping only their own notes, and say what they make of it. A
+comparer that never saw the plan marks their words against the track.
+`scripts/cold_watch.py` writes every call, checks the marks and says whether the floor is met:
+every must-know hit in time by two of the three, and no wrong ending. It is the one check of
+whether a stranger follows the video: nobody who knows the plan, you included, can judge that.
+Its viewer calls cost credits unless their quote shows 0 for every one: quote them as
+`references/cold-watch.md` says and 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 the first round; a credit or storage error
+stops it, never retried. It is not a measure of retention, and it says so.
 
-**The critic panel.** After the viewer panel, before the user sees a draft, six simulated critics
+**The critic panel.** After the cold watch, before the user sees a draft, six simulated critics
 with taste each look at it through one lens: the first-principles critic (the core problem the
 video solves), the radical (the concept), the interface designer (the UI on screen), the story
 editor (the message and the script), the growth strategist (the stop, the drops, the ending) and
 the art director (the look). Each sees only frames every 0.5 s, the source stills (the real UI or
-footage the video depicts), the words, the goal and the destination, and returns a verdict, problems with their times, and fixes. `scripts/critics.py`
+footage the video depicts), the words, the goal, the destination and the cold watch's gap list, and returns a verdict, problems with their times, and fixes. `scripts/critics.py`
 picks the frames and lays the notes side by side. Where critics agree it is a must-fix, weighed
 against the first-principles critic's core sentence; a new concept goes to the user as a choice,
-never applied silently. A draft that looks different from its source is a must-fix.
+never applied silently. So does a shot a critic would swap in an edit of the user's footage: it
+is a proposal, shown to the user as a rendered A/B, never a silent change to the plan. A draft that looks different from its source is a must-fix.
 
 **The challenger.** A seventh critic who does not wait for the draft. At each decision the use case
 names (the idea, the plan, the key frames and each image picked, the UI, and every SPEND before
 its quote) it asks "why this?" and tries to show that the direction is wrong before more work is
 spent on it. It returns `keep`, `rethink` or `stop`, with the reason and the cheaper alternative.
 It also gets the source stills, so a design that drifts from the real app is caught as it is made.
+Its `rethink` of a shot the user saw and kept is a proposal too: an A/B, never a silent swap.
+
+**The QC inspector.** An edit of the user's own footage is checked from its render by an
+inspector who did not make it and never sees why it was made that way: camera steadiness on every
+used range, lips against the voice, each cut against its word, shot lengths, legible text,
+captions, repeated or clipped lines, the locked meaning map, and every keeper of a version the
+user saw: one dropped with no line the user will read fails the draft, and, when a reference's
+format is in the record, the reference match (`scripts/reference_match.py` reads it from the
+scene). It reports only critical and
+major findings, each with its method and confidence, and marks what it could not see or hear as
+open, never as a pass. `scripts/inspection.py` reads its report back.
 
 | Read this | Before you |
 |---|---|
-| `references/viewer-panel.md` | run the viewer panel on a hook or a draft |
+| `references/qc-inspector.md` | check a draft of an edit of the user's footage, or a fix to one |
+| `references/cold-watch.md` | write a viewer track, or run the cold watch on a draft |
+| `references/cold-watch-comparer.md` | mark a cold watch's accounts against the track (the comparer's prompt) |
 | `references/critic-panel.md` | run the critic panel on a draft |
 | `references/challenger.md` | go on from a decision the use case names, or show a quote |
 
Full technical diff · 3 changed fields

changed /description

BEFORE
"Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a short's script, hook and draft reviewed by a simulated viewer panel and critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names."
AFTER
"Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names."

changed /included_files

BEFORE
[
  {
    "relative_path": "references/agent-challenger.md",
    "size_in_bytes": 3399
  },
  {
    "relative_path": "references/agent-critic-art.md",
    "size_in_bytes": 3507
  },
  {
    "relative_path": "references/agent-critic-growth.md",
    "size_in_bytes": 3688
  },
  {
    "relative_path": "references/agent-critic-interface.md",
    "size_in_bytes": 3858
  },
  {
    "relative_path": "references/agent-critic-radical.md",
    "size_in_bytes": 3615
  },
  {
    "relative_path": "references/agent-critic-roots.md",
    "size_in_bytes": 4467
  },
  {
    "relative_path": "references/agent-critic-story.md",
    "size_in_bytes": 3675
  },
  {
    "relative_path": "references/challenger.md",
    "size_in_bytes": 3825
  },
  {
    "relative_path": "references/critic-panel.md",
    "size_in_bytes": 8145
  },
  {
    "relative_path": "references/readiness-verdict.md",
    "size_in_bytes": 3340
  },
  {
    "relative_path": "references/script-review.md",
    "size_in_bytes": 3590
  },
  {
    "relative_path": "references/viewer-panel.md",
    "size_in_bytes": 5199
  },
  {
    "relative_path": "scripts/checklist.py",
    "size_in_bytes": 15761
  },
  {
    "relative_path": "scripts/cover.py",
    "size_in_bytes": 16819
  },
  {
    "relative_path": "scripts/critics.py",
    "size_in_bytes": 12986
  },
  {
    "relative_path": "scripts/cut_check.py",
    "size_in_bytes": 57815
  },
  {
    "relative_path": "scripts/framing.py",
    "size_in_bytes": 7644
  },
  {
    "relative_path": "scripts/panel.py",
    "size_in_bytes": 13985
  },
  {
    "relative_path": "scripts/panels.py",
    "size_in_bytes": 13591
  },
  {
    "relative_path": "scripts/qc.py",
    "size_in_bytes": 14424
  },
  {
    "relative_path": "scripts/readability.py",
    "size_in_bytes": 7792
  },
  {
    "relative_path": "scripts/safe_zone.py",
    "size_in_bytes": 61049
  },
  {
    "relative_path": "scripts/script_review.py",
    "size_in_bytes": 26893
  },
  {
    "relative_path": "scripts/timing.py",
    "size_in_bytes": 32167
  }
]
AFTER
[
  {
    "relative_path": "references/agent-challenger.md",
    "size_in_bytes": 3399
  },
  {
    "relative_path": "references/agent-critic-art.md",
    "size_in_bytes": 3825
  },
  {
    "relative_path": "references/agent-critic-growth.md",
    "size_in_bytes": 4201
  },
  {
    "relative_path": "references/agent-critic-interface.md",
    "size_in_bytes": 4176
  },
  {
    "relative_path": "references/agent-critic-radical.md",
    "size_in_bytes": 3933
  },
  {
    "relative_path": "references/agent-critic-roots.md",
    "size_in_bytes": 4785
  },
  {
    "relative_path": "references/agent-critic-story.md",
    "size_in_bytes": 4180
  },
  {
    "relative_path": "references/agent-qc-inspector.md",
    "size_in_bytes": 5917
  },
  {
    "relative_path": "references/challenger.md",
    "size_in_bytes": 3825
  },
  {
    "relative_path": "references/cold-watch-comparer.md",
    "size_in_bytes": 2832
  },
  {
    "relative_path": "references/cold-watch.md",
    "size_in_bytes": 13288
  },
  {
    "relative_path": "references/critic-panel.md",
    "size_in_bytes": 8774
  },
  {
    "relative_path": "references/qc-inspector.md",
    "size_in_bytes": 3992
  },
  {
    "relative_path": "references/readiness-verdict.md",
    "size_in_bytes": 3340
  },
  {
    "relative_path": "references/script-review.md",
    "size_in_bytes": 3590
  },
  {
    "relative_path": "scripts/_words.py",
    "size_in_bytes": 3680
  },
  {
    "relative_path": "scripts/checklist.py",
    "size_in_bytes": 15762
  },
  {
    "relative_path": "scripts/cold_watch.py",
    "size_in_bytes": 56571
  },
  {
    "relative_path": "scripts/cover.py",
    "size_in_bytes": 16819
  },
  {
    "relative_path": "scripts/critics.py",
    "size_in_bytes": 12987
  },
  {
    "relative_path": "scripts/cut_check.py",
    "size_in_bytes": 113194
  },
  {
    "relative_path": "scripts/framing.py",
    "size_in_bytes": 7644
  },
  {
    "relative_path": "scripts/inspection.py",
    "size_in_bytes": 5334
  },
  {
    "relative_path": "scripts/panels.py",
    "size_in_bytes": 13591
  },
  {
    "relative_path": "scripts/qc.py",
    "size_in_bytes": 14424
  },
  {
    "relative_path": "scripts/readability.py",
    "size_in_bytes": 7792
  },
  {
    "relative_path": "scripts/reference_match.py",
    "size_in_bytes": 19553
  },
  {
    "relative_path": "scripts/safe_zone.py",
    "size_in_bytes": 61049
  },
  {
    "relative_path": "scripts/script_review.py",
    "size_in_bytes": 26893
  },
  {
    "relative_path": "scripts/timing.py",
    "size_in_bytes": 32913
  }
]

changed /skill_md_contents

BEFORE
"---\nname: quality-review\ndescription: Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a short's script, hook and draft reviewed by a simulated viewer panel and critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.\nmetadata:\n  imports: screen-recording-layouts\n---\n\n# Evaluate\n\n1. **Validate.** A successful `picsart_scene_patch` has already validated its result, so call\n   `picsart_scene_validate` only after changes made any other way. Zero errors; fix the\n   warnings too.\n2. **Lint the layout.** `picsart_scene_lint_layout` over the timeline; on a long scene, narrow\n   `t0`/`t1` to the part in question. Each of its `findings` is a sustained overlap between layers\n   that should not touch. For a pair placed together on purpose, pass `allow: [\"idA:idB\"]`. Leave\n   `engine` at its default (`native`), which measures real text boxes. A clean lint is not proof:\n   on the `jet` engine it never measures text (a fallback), and on `v3` a text box is the font's\n   line box, so it can flag empty space. If the tool is not in your list, or fails because the\n   query build is unavailable, skip to the frames.\n3. **Measure.** For a flagged pair, or a use-case question about where something sits (inside a\n   panel, inside a mask), `picsart_scene_query_layout` at that moment gives each layer's box. It\n   reads layer geometry, not painted pixels, and has no box for text on `jet`.\n4. `picsart_scene_contact_sheet` at each moment the use case names. Look at every frame and answer\n   the use case's questions about it.\n5. `picsart_scene_contact_sheet` once across the whole timeline — order, seams, nothing blank.\n   Then, for every cut, its `times` at the cut and one frame either side (`cut − 1/fps`, `cut`,\n   `cut + 1/fps`): a one-frame black gap falls between evenly spaced samples.\n6. **Polish**, once the frames are right. A correct video can still look machine-made, and the\n   user notices that first:\n   - when the use case built it in a style, run the visual-style skill's `adherence.py` on it,\n     and check the frames against that skill's list of defaults that look machine-made;\n   - **the first frame is a real moment**: the payoff, or the problem it solves, as the\n     retention-hooks skill's hook rules say. Not a login screen, an empty app or a title on a plain\n     canvas;\n   - **every second has something new**: no stretch over 3 s where nothing moves, appears or\n     changes, unless it is a hold the use case asked for. First run `scripts/timing.py` on\n     `{ scene }`, from this skill's folder: from the JSON alone it flags every stretch over 8 s\n     with no change (`dead_air`), every block of words held under 0.3 s a word or 1 s\n     (`short_hold`), and every block over 10 words, or a step label over 6 (`too_many_words`).\n     Typed holds add only the typing remaining in each visible span, including after retiming or\n     trimming. Text and caption masks in `mask.clip` or legacy `mask.layer` need rendered review;\n     the script rejects them.\n     Pass the measured reveals inside a clip as `changes`; the JSON cannot see them. Then look\n     at the frames for the 3 s: motion inside a clip is only on them;\n   - **not one move everywhere**: when every beat has the same push or the same entrance, vary\n     what the use case's skills allow;\n   - **words inside the safe zone** of the destination the use case names: run the screen-recording-layouts skill's\n     `safe_zone.py` with its profile, and fix each finding. Readable at the frame's width.\n\n   Fix each finding, or keep it and say why in the review line (\"the app itself is navy, so the\n   canvas follows it\").\n7. Fix the layers and check again. Building is free.\n\n**The brief, item by item.** When the use case has a brief (the message, the destinations, the\nlength, the hook, the call to action, the steps), grade the build against it one item at a time,\nnever as one judgement of the whole. From this skill's folder, run\n`python3 scripts/checklist.py <file>` on `{ brief, scene }`, with the destinations as the screen-recording-layouts\nskill's safe-zone profiles. It turns the brief into yes-or-no questions and grades the ones the\nscene answers: the length, words by 0.5 s, the hook's words, the call to action at the end, the\nsafe zone, and `timing.py`. Checks the scene cannot answer for a text mask become `look` items,\nalongside the remaining checks. Answer each from the frames of steps 4-6, with the frame's\ntime or the tool's line as `evidence`, and run it again with your `answers`; an answer without\nevidence stays open. Fix only the items that say no, then grade again: never rebuild what passed.\nIts `summary` goes with the review.\n\n**The viewer panel.** On a short, before the user sees its hook or its draft, three simulated\nviewers (a 15-year-old scrolling, a creator new to the topic, someone who already uses the\nproduct) look at the first 2 s and then at the whole draft. Their answers are graded against a\npass bar with `scripts/panel.py`. It catches a hook that reads as an ad or promises the wrong\nthing. It is not a measure of retention, and it says so.\n\n**The critic panel.** After the viewer panel, before the user sees a draft, six simulated critics\nwith taste each look at it through one lens: the first-principles critic (the core problem the\nvideo solves), the radical (the concept), the interface designer (the UI on screen), the story\neditor (the message and the script), the growth strategist (the stop, the drops, the ending) and\nthe art director (the look). Each sees only frames every 0.5 s, the source stills (the real UI or\nfootage the video depicts), the words, the goal and the destination, and returns a verdict, problems with their times, and fixes. `scripts/critics.py`\npicks the frames and lays the notes side by side. Where critics agree it is a must-fix, weighed\nagainst the first-principles critic's core sentence; a new concept goes to the user as a choice,\nnever applied silently. A draft that looks different from its source is a must-fix.\n\n**The challenger.** A seventh critic who does not wait for the draft. At each decision the use case\nnames (the idea, the plan, the key frames and each image picked, the UI, and every SPEND before\nits quote) it asks \"why this?\" and tries to show that the direction is wrong before more work is\nspent on it. It returns `keep`, `rethink` or `stop`, with the reason and the cheaper alternative.\nIt also gets the source stills, so a design that drifts from the real app is caught as it is made.\n\n| Read this | Before you |\n|---|---|\n| `references/viewer-panel.md` | run the viewer panel on a hook or a draft |\n| `references/critic-panel.md` | run the critic panel on a draft |\n| `references/challenger.md` | go on from a decision the use case names, or show a quote |\n\n**When a frame or an export fails to render.** The scene validated, so the fault is in a feature\nthe renderer refuses, and the error names it rarely. Find it before changing anything:\n\n1. Copy the error text exactly.\n2. Cut a copy of the scene down to the one layer at the failing time, and sample that. Then take\n   away one feature at a time (its mask, its effects, its animation) until it renders. Each sample is\n   one frame.\n3. Keep the scene as it is. Tell the user which feature fails, with the error and the smallest\n   scene that shows it, and offer a way round it that keeps the look where one exists (for a mask,\n   the screen-recording-layouts skill's nested panel). Never reshape the\n   whole scene to guess past it.\n\n**Before a film's cut is reviewed.** From this skill's folder, run\n`python3 scripts/cut_check.py <file>` on `{ scene, allowed?, shots?, places? }`: the cut's scene,\nits allowed spans (the screen clips placed in it, as film-cutting says), and from the project record\nthe shot each clip shows (keyed by clip id, with its size and angle when the shot list has them) and\nthe place each shows. From the scene alone it finds the joins, moves and overlays a viewer reads\nas a glitch: two takes of one shot, or two shots of one size and angle, cut together\n(`same-scene-cut`), or a cut hidden under a quick move (`hidden-take-change`); different shots of\none place are coverage and pass, a short dissolve between takes (`ghost`), a move of the picture, or of a layer covering 40% of the frame or more, that starts or stops in\none frame (`pop-move`), an overlay gone for a frame or two (`blink`), a zoom over a still picture\n(`zoom-idle`), and a screen held still inside its span (`screen-still`). Nothing inside a span is\nflagged except `screen-still`. It prints `{ ok, errors, warnings, unchecked, skipped }`: fix each error as\nfilm-cutting says and run it again, say each warning in the review, and check what it lists as\n`unchecked` on the frames of step 5, never calling it checked. On an animatic of held stills, add\n`--slides`: its stills hold on purpose, so the motion rules move to `skipped` and run on the cut,\nand the joins, overlays and screens are still checked.\n\n**Before you hand over an export.** An export can differ from its scene: another size, a dropped\nframe rate, a missing sound track, a silent last second. Probe the file with\n`picsart_media_probe`, which is free, and from this skill's folder run\n`python3 scripts/qc.py <file>` on `{ scene, probe, audio? }`: the scene you exported, the probe as\nit returned, and in `audio` only what a tool measured (loudness, true peak, the sound's length, an\nenergy curve). It prints `{ ok, errors, warnings, unchecked }`:\n\n- an `error` (the size, the fps or the length is not the scene's, the sound is missing, drifts or\n  peaks above -0.5 dBTP): make its `fix` and export again, once. Still there: deliver, and name it;\n- a `warning` (loudness away from -14 LUFS, a silent tail over 1.2 s): say it in the delivery\n  line;\n- `unchecked`: what no tool measured. Never call it checked.\n\nFrames show framing, not motion: the user judges timing and easing in REVIEW.\n\nHow many rounds you get, and what to do when they run out, is in picsart-video-rules (Evaluation). Never\nexport to \"show\" a draft that the user can review live; picsart-video-rules (Preview and export) says when\nto export.\n\n## The script, before the build\n\nBefore a short is built, voiced or generated, review its script: is it worth making, what is its\nwhy, and which lines to rewrite. A video whose footage never shows its result is dropped before it\ncosts anything. The review runs `scripts/script_review.py`; the reference below has its steps.\n\n| Read this | Before you |\n|---|---|\n| `references/script-review.md` | review a script, or report its verdict |\n| `references/readiness-verdict.md` | declare a generated video's decisive evidence before paying for it, or say whether a candidate is ready |\n"
AFTER
"---\nname: quality-review\ndescription: Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.\nmetadata:\n  imports: screen-recording-layouts\n---\n\n# Evaluate\n\n1. **Validate.** A successful `picsart_scene_patch` has already validated its result, so call\n   `picsart_scene_validate` only after changes made any other way. Zero errors; fix the\n   warnings too.\n2. **Lint the layout.** `picsart_scene_lint_layout` over the timeline; on a long scene, narrow\n   `t0`/`t1` to the part in question. Each of its `findings` is a sustained overlap between layers\n   that should not touch. For a pair placed together on purpose, pass `allow: [\"idA:idB\"]`. A text\n   box is the font's line box, so a clean lint is not proof and it can flag empty space. If the tool is not in your list, or fails because the\n   query build is unavailable, skip to the frames.\n3. **Measure.** For a flagged pair, or a use-case question about where something sits (inside a\n   panel, inside a mask), `picsart_scene_query_layout` at that moment gives each layer's box. It\n   reads layer geometry, not painted pixels.\n4. `picsart_scene_contact_sheet` at each moment the use case names. Look at every frame and answer\n   the use case's questions about it.\n5. `picsart_scene_contact_sheet` once across the whole timeline — order, seams, nothing blank.\n   Then, for every cut, its `times` at the cut and one frame either side (`cut − 1/fps`, `cut`,\n   `cut + 1/fps`): a one-frame black gap falls between evenly spaced samples.\n6. **Polish**, once the frames are right. A correct video can still look machine-made, and the\n   user notices that first:\n   - when the use case built it in a style, run the visual-style skill's `adherence.py` on it,\n     and check the frames against that skill's list of defaults that look machine-made;\n   - **the first frame is a real moment**: the payoff, or the problem it solves, as the\n     retention-hooks skill's hook rules say. Not a login screen, an empty app or a title on a plain\n     canvas;\n   - **every second has something new**: no stretch over 3 s where nothing moves, appears or\n     changes, unless it is a hold the use case asked for. First run `scripts/timing.py` on\n     `{ scene }`, from this skill's folder: from the JSON alone it flags every stretch over 8 s\n     with no change (`dead_air`), every block of words held under 0.3 s a word or 1 s\n     (`short_hold`), and every block over 10 words, or a step label over 6 (`too_many_words`).\n     Typed holds add only the typing remaining in each visible span, including after retiming or\n     trimming. Text and caption masks in `mask.clip` or legacy `mask.layer` need rendered review;\n     the script rejects them.\n     Pass the measured reveals inside a clip as `changes`; the JSON cannot see them. Then look\n     at the frames for the 3 s: motion inside a clip is only on them;\n   - **not one move everywhere**: when every beat has the same push or the same entrance, vary\n     what the use case's skills allow;\n   - **words inside the safe zone** of the destination the use case names: run the screen-recording-layouts skill's\n     `safe_zone.py` with its profile, and fix each finding. Readable at the frame's width.\n\n   Fix each finding, or keep it and say why in the review line (\"the app itself is navy, so the\n   canvas follows it\").\n7. Fix the layers and check again. Building is free.\n\n**The brief, item by item.** When the use case has a brief (the message, the destinations, the\nlength, the hook, the call to action, the steps), grade the build against it one item at a time,\nnever as one judgement of the whole. From this skill's folder, run\n`python3 scripts/checklist.py <file>` on `{ brief, scene }`, with the destinations as the screen-recording-layouts\nskill's safe-zone profiles. It turns the brief into yes-or-no questions and grades the ones the\nscene answers: the length, words by 0.5 s, the hook's words, the call to action at the end, the\nsafe zone, and `timing.py`. Checks the scene cannot answer for a text mask become `look` items,\nalongside the remaining checks. Answer each from the frames of steps 4-6, with the frame's\ntime or the tool's line as `evidence`, and run it again with your `answers`; an answer without\nevidence stays open. Fix only the items that say no, then grade again: never rebuild what passed.\nIts `summary` goes with the review.\n\n**The cold watch.** Before the critics, and before the user sees a draft that has a viewer track,\nthree model viewers who know nothing of it watch it forward, a few seconds at a time, from its\nframes and its heard words, each keeping only their own notes, and say what they make of it. A\ncomparer that never saw the plan marks their words against the track.\n`scripts/cold_watch.py` writes every call, checks the marks and says whether the floor is met:\nevery must-know hit in time by two of the three, and no wrong ending. It is the one check of\nwhether a stranger follows the video: nobody who knows the plan, you included, can judge that.\nIts viewer calls cost credits unless their quote shows 0 for every one: quote them as\n`references/cold-watch.md` says and 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 the first round; a credit or storage error\nstops it, never retried. It is not a measure of retention, and it says so.\n\n**The critic panel.** After the cold watch, before the user sees a draft, six simulated critics\nwith taste each look at it through one lens: the first-principles critic (the core problem the\nvideo solves), the radical (the concept), the interface designer (the UI on screen), the story\neditor (the message and the script), the growth strategist (the stop, the drops, the ending) and\nthe art director (the look). Each sees only frames every 0.5 s, the source stills (the real UI or\nfootage the video depicts), the words, the goal, the destination and the cold watch's gap list, and returns a verdict, problems with their times, and fixes. `scripts/critics.py`\npicks the frames and lays the notes side by side. Where critics agree it is a must-fix, weighed\nagainst the first-principles critic's core sentence; a new concept goes to the user as a choice,\nnever applied silently. So does a shot a critic would swap in an edit of the user's footage: it\nis a proposal, shown to the user as a rendered A/B, never a silent change to the plan. A draft that looks different from its source is a must-fix.\n\n**The challenger.** A seventh critic who does not wait for the draft. At each decision the use case\nnames (the idea, the plan, the key frames and each image picked, the UI, and every SPEND before\nits quote) it asks \"why this?\" and tries to show that the direction is wrong before more work is\nspent on it. It returns `keep`, `rethink` or `stop`, with the reason and the cheaper alternative.\nIt also gets the source stills, so a design that drifts from the real app is caught as it is made.\nIts `rethink` of a shot the user saw and kept is a proposal too: an A/B, never a silent swap.\n\n**The QC inspector.** An edit of the user's own footage is checked from its render by an\ninspector who did not make it and never sees why it was made that way: camera steadiness on every\nused range, lips against the voice, each cut against its word, shot lengths, legible text,\ncaptions, repeated or clipped lines, the locked meaning map, and every keeper of a version the\nuser saw: one dropped with no line the user will read fails the draft, and, when a reference's\nformat is in the record, the reference match (`scripts/reference_match.py` reads it from the\nscene). It reports only critical and\nmajor findings, each with its method and confidence, and marks what it could not see or hear as\nopen, never as a pass. `scripts/inspection.py` reads its report back.\n\n| Read this | Before you |\n|---|---|\n| `references/qc-inspector.md` | check a draft of an edit of the user's footage, or a fix to one |\n| `references/cold-watch.md` | write a viewer track, or run the cold watch on a draft |\n| `references/cold-watch-comparer.md` | mark a cold watch's accounts against the track (the comparer's prompt) |\n| `references/critic-panel.md` | run the critic panel on a draft |\n| `references/challenger.md` | go on from a decision the use case names, or show a quote |\n\n**When a frame or an export fails to render.** The scene validated, so the fault is in a feature\nthe renderer refuses, and the error names it rarely. Find it before changing anything:\n\n1. Copy the error text exactly.\n2. Cut a copy of the scene down to the one layer at the failing time, and sample that. Then take\n   away one feature at a time (its mask, its effects, its animation) until it renders. Each sample is\n   one frame.\n3. Keep the scene as it is. Tell the user which feature fails, with the error and the smallest\n   scene that shows it, and offer a way round it that keeps the look where one exists (for a mask,\n   the screen-recording-layouts skill's nested panel). Never reshape the\n   whole scene to guess past it.\n\n**Before a film's cut is reviewed.** From this skill's folder, run\n`python3 scripts/cut_check.py <file>` on `{ scene, allowed?, shots?, places? }`: the cut's scene,\nits allowed spans (the screen clips placed in it, as film-cutting says), and from the project record\nthe shot each clip shows (keyed by clip id, with its size and angle when the shot list has them) and\nthe place each shows. From the scene alone it finds the joins, moves and overlays a viewer reads\nas a glitch: two takes of one shot, or two shots of one size and angle, cut together\n(`same-scene-cut`), or a cut hidden under a quick move (`hidden-take-change`); different shots of\none place are coverage and pass, a short dissolve between takes (`ghost`), a move of the picture, or of a layer covering 40% of the frame or more, that starts or stops in\none frame (`pop-move`), an overlay gone for a frame or two (`blink`), a zoom over a still picture\n(`zoom-idle`), and a screen held still inside its span (`screen-still`). Nothing inside a span is\nflagged except `screen-still`. It prints `{ ok, errors, warnings, unchecked, skipped }`: fix each error as\nfilm-cutting says and run it again, say each warning in the review, and check what it lists as\n`unchecked` on the frames of step 5, never calling it checked. On an animatic of held stills, add\n`--slides`: its stills hold on purpose, so the motion rules move to `skipped` and run on the cut,\nand the joins, overlays and screens are still checked.\n\n**Before you hand over an export.** An export can differ from its scene: another size, a dropped\nframe rate, a missing sound track, a silent last second. Probe the file with\n`picsart_media_probe`, which is free, and from this skill's folder run\n`python3 scripts/qc.py <file>` on `{ scene, probe, audio? }`: the scene you exported, the probe as\nit returned, and in `audio` only what a tool measured (loudness, true peak, the sound's length, an\nenergy curve). It prints `{ ok, errors, warnings, unchecked }`:\n\n- an `error` (the size, the fps or the length is not the scene's, the sound is missing, drifts or\n  peaks above -0.5 dBTP): make its `fix` and export again, once. Still there: deliver, and name it;\n- a `warning` (loudness away from -14 LUFS, a silent tail over 1.2 s): say it in the delivery\n  line;\n- `unchecked`: what no tool measured. Never call it checked.\n\nFrames show framing, not motion: the user judges timing and easing in REVIEW.\n\nHow many rounds you get, and what to do when they run out, is in picsart-video-rules (Evaluation). Never\nexport to \"show\" a draft that the user can review live; picsart-video-rules (Preview and export) says when\nto export.\n\n## The script, before the build\n\nBefore a short is built, voiced or generated, review its script: is it worth making, what is its\nwhy, and which lines to rewrite. A video whose footage never shows its result is dropped before it\ncosts anything. The review runs `scripts/script_review.py`; the reference below has its steps.\n\n| Read this | Before you |\n|---|---|\n| `references/script-review.md` | review a script, or report its verdict |\n| `references/readiness-verdict.md` | declare a generated video's decisive evidence before paying for it, or say whether a candidate is ready |\n"
Full snapshot data
{
  "description": "Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.",
  "included_files": [
    {
      "relative_path": "references/agent-challenger.md",
      "size_in_bytes": 3399
    },
    {
      "relative_path": "references/agent-critic-art.md",
      "size_in_bytes": 3825
    },
    {
      "relative_path": "references/agent-critic-growth.md",
      "size_in_bytes": 4201
    },
    {
      "relative_path": "references/agent-critic-interface.md",
      "size_in_bytes": 4176
    },
    {
      "relative_path": "references/agent-critic-radical.md",
      "size_in_bytes": 3933
    },
    {
      "relative_path": "references/agent-critic-roots.md",
      "size_in_bytes": 4785
    },
    {
      "relative_path": "references/agent-critic-story.md",
      "size_in_bytes": 4180
    },
    {
      "relative_path": "references/agent-qc-inspector.md",
      "size_in_bytes": 5917
    },
    {
      "relative_path": "references/challenger.md",
      "size_in_bytes": 3825
    },
    {
      "relative_path": "references/cold-watch-comparer.md",
      "size_in_bytes": 2832
    },
    {
      "relative_path": "references/cold-watch.md",
      "size_in_bytes": 13288
    },
    {
      "relative_path": "references/critic-panel.md",
      "size_in_bytes": 8774
    },
    {
      "relative_path": "references/qc-inspector.md",
      "size_in_bytes": 3992
    },
    {
      "relative_path": "references/readiness-verdict.md",
      "size_in_bytes": 3340
    },
    {
      "relative_path": "references/script-review.md",
      "size_in_bytes": 3590
    },
    {
      "relative_path": "scripts/_words.py",
      "size_in_bytes": 3680
    },
    {
      "relative_path": "scripts/checklist.py",
      "size_in_bytes": 15762
    },
    {
      "relative_path": "scripts/cold_watch.py",
      "size_in_bytes": 56571
    },
    {
      "relative_path": "scripts/cover.py",
      "size_in_bytes": 16819
    },
    {
      "relative_path": "scripts/critics.py",
      "size_in_bytes": 12987
    },
    {
      "relative_path": "scripts/cut_check.py",
      "size_in_bytes": 113194
    },
    {
      "relative_path": "scripts/framing.py",
      "size_in_bytes": 7644
    },
    {
      "relative_path": "scripts/inspection.py",
      "size_in_bytes": 5334
    },
    {
      "relative_path": "scripts/panels.py",
      "size_in_bytes": 13591
    },
    {
      "relative_path": "scripts/qc.py",
      "size_in_bytes": 14424
    },
    {
      "relative_path": "scripts/readability.py",
      "size_in_bytes": 7792
    },
    {
      "relative_path": "scripts/reference_match.py",
      "size_in_bytes": 19553
    },
    {
      "relative_path": "scripts/safe_zone.py",
      "size_in_bytes": 61049
    },
    {
      "relative_path": "scripts/script_review.py",
      "size_in_bytes": 26893
    },
    {
      "relative_path": "scripts/timing.py",
      "size_in_bytes": 32913
    }
  ],
  "name": "quality-review",
  "skill_md_contents": "---\nname: quality-review\ndescription: Check work before the user sees it - a scene validated, linted and looked at in contact-sheet frames, then fixed and checked again; a draft watched cold by model viewers against its viewer track, then reviewed by critics; the brief graded; an export checked against its scene. Use after a script is written and before it is built or paid for, after building or changing a scene, before REVIEW or an export, before handing over an export, and at each decision a use case names.\nmetadata:\n  imports: screen-recording-layouts\n---\n\n# Evaluate\n\n1. **Validate.** A successful `picsart_scene_patch` has already validated its result, so call\n   `picsart_scene_validate` only after changes made any other way. Zero errors; fix the\n   warnings too.\n2. **Lint the layout.** `picsart_scene_lint_layout` over the timeline; on a long scene, narrow\n   `t0`/`t1` to the part in question. Each of its `findings` is a sustained overlap between layers\n   that should not touch. For a pair placed together on purpose, pass `allow: [\"idA:idB\"]`. A text\n   box is the font's line box, so a clean lint is not proof and it can flag empty space. If the tool is not in your list, or fails because the\n   query build is unavailable, skip to the frames.\n3. **Measure.** For a flagged pair, or a use-case question about where something sits (inside a\n   panel, inside a mask), `picsart_scene_query_layout` at that moment gives each layer's box. It\n   reads layer geometry, not painted pixels.\n4. `picsart_scene_contact_sheet` at each moment the use case names. Look at every frame and answer\n   the use case's questions about it.\n5. `picsart_scene_contact_sheet` once across the whole timeline — order, seams, nothing blank.\n   Then, for every cut, its `times` at the cut and one frame either side (`cut − 1/fps`, `cut`,\n   `cut + 1/fps`): a one-frame black gap falls between evenly spaced samples.\n6. **Polish**, once the frames are right. A correct video can still look machine-made, and the\n   user notices that first:\n   - when the use case built it in a style, run the visual-style skill's `adherence.py` on it,\n     and check the frames against that skill's list of defaults that look machine-made;\n   - **the first frame is a real moment**: the payoff, or the problem it solves, as the\n     retention-hooks skill's hook rules say. Not a login screen, an empty app or a title on a plain\n     canvas;\n   - **every second has something new**: no stretch over 3 s where nothing moves, appears or\n     changes, unless it is a hold the use case asked for. First run `scripts/timing.py` on\n     `{ scene }`, from this skill's folder: from the JSON alone it flags every stretch over 8 s\n     with no change (`dead_air`), every block of words held under 0.3 s a word or 1 s\n     (`short_hold`), and every block over 10 words, or a step label over 6 (`too_many_words`).\n     Typed holds add only the typing remaining in each visible span, including after retiming or\n     trimming. Text and caption masks in `mask.clip` or legacy `mask.layer` need rendered review;\n     the script rejects them.\n     Pass the measured reveals inside a clip as `changes`; the JSON cannot see them. Then look\n     at the frames for the 3 s: motion inside a clip is only on them;\n   - **not one move everywhere**: when every beat has the same push or the same entrance, vary\n     what the use case's skills allow;\n   - **words inside the safe zone** of the destination the use case names: run the screen-recording-layouts skill's\n     `safe_zone.py` with its profile, and fix each finding. Readable at the frame's width.\n\n   Fix each finding, or keep it and say why in the review line (\"the app itself is navy, so the\n   canvas follows it\").\n7. Fix the layers and check again. Building is free.\n\n**The brief, item by item.** When the use case has a brief (the message, the destinations, the\nlength, the hook, the call to action, the steps), grade the build against it one item at a time,\nnever as one judgement of the whole. From this skill's folder, run\n`python3 scripts/checklist.py <file>` on `{ brief, scene }`, with the destinations as the screen-recording-layouts\nskill's safe-zone profiles. It turns the brief into yes-or-no questions and grades the ones the\nscene answers: the length, words by 0.5 s, the hook's words, the call to action at the end, the\nsafe zone, and `timing.py`. Checks the scene cannot answer for a text mask become `look` items,\nalongside the remaining checks. Answer each from the frames of steps 4-6, with the frame's\ntime or the tool's line as `evidence`, and run it again with your `answers`; an answer without\nevidence stays open. Fix only the items that say no, then grade again: never rebuild what passed.\nIts `summary` goes with the review.\n\n**The cold watch.** Before the critics, and before the user sees a draft that has a viewer track,\nthree model viewers who know nothing of it watch it forward, a few seconds at a time, from its\nframes and its heard words, each keeping only their own notes, and say what they make of it. A\ncomparer that never saw the plan marks their words against the track.\n`scripts/cold_watch.py` writes every call, checks the marks and says whether the floor is met:\nevery must-know hit in time by two of the three, and no wrong ending. It is the one check of\nwhether a stranger follows the video: nobody who knows the plan, you included, can judge that.\nIts viewer calls cost credits unless their quote shows 0 for every one: quote them as\n`references/cold-watch.md` says and 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 the first round; a credit or storage error\nstops it, never retried. It is not a measure of retention, and it says so.\n\n**The critic panel.** After the cold watch, before the user sees a draft, six simulated critics\nwith taste each look at it through one lens: the first-principles critic (the core problem the\nvideo solves), the radical (the concept), the interface designer (the UI on screen), the story\neditor (the message and the script), the growth strategist (the stop, the drops, the ending) and\nthe art director (the look). Each sees only frames every 0.5 s, the source stills (the real UI or\nfootage the video depicts), the words, the goal, the destination and the cold watch's gap list, and returns a verdict, problems with their times, and fixes. `scripts/critics.py`\npicks the frames and lays the notes side by side. Where critics agree it is a must-fix, weighed\nagainst the first-principles critic's core sentence; a new concept goes to the user as a choice,\nnever applied silently. So does a shot a critic would swap in an edit of the user's footage: it\nis a proposal, shown to the user as a rendered A/B, never a silent change to the plan. A draft that looks different from its source is a must-fix.\n\n**The challenger.** A seventh critic who does not wait for the draft. At each decision the use case\nnames (the idea, the plan, the key frames and each image picked, the UI, and every SPEND before\nits quote) it asks \"why this?\" and tries to show that the direction is wrong before more work is\nspent on it. It returns `keep`, `rethink` or `stop`, with the reason and the cheaper alternative.\nIt also gets the source stills, so a design that drifts from the real app is caught as it is made.\nIts `rethink` of a shot the user saw and kept is a proposal too: an A/B, never a silent swap.\n\n**The QC inspector.** An edit of the user's own footage is checked from its render by an\ninspector who did not make it and never sees why it was made that way: camera steadiness on every\nused range, lips against the voice, each cut against its word, shot lengths, legible text,\ncaptions, repeated or clipped lines, the locked meaning map, and every keeper of a version the\nuser saw: one dropped with no line the user will read fails the draft, and, when a reference's\nformat is in the record, the reference match (`scripts/reference_match.py` reads it from the\nscene). It reports only critical and\nmajor findings, each with its method and confidence, and marks what it could not see or hear as\nopen, never as a pass. `scripts/inspection.py` reads its report back.\n\n| Read this | Before you |\n|---|---|\n| `references/qc-inspector.md` | check a draft of an edit of the user's footage, or a fix to one |\n| `references/cold-watch.md` | write a viewer track, or run the cold watch on a draft |\n| `references/cold-watch-comparer.md` | mark a cold watch's accounts against the track (the comparer's prompt) |\n| `references/critic-panel.md` | run the critic panel on a draft |\n| `references/challenger.md` | go on from a decision the use case names, or show a quote |\n\n**When a frame or an export fails to render.** The scene validated, so the fault is in a feature\nthe renderer refuses, and the error names it rarely. Find it before changing anything:\n\n1. Copy the error text exactly.\n2. Cut a copy of the scene down to the one layer at the failing time, and sample that. Then take\n   away one feature at a time (its mask, its effects, its animation) until it renders. Each sample is\n   one frame.\n3. Keep the scene as it is. Tell the user which feature fails, with the error and the smallest\n   scene that shows it, and offer a way round it that keeps the look where one exists (for a mask,\n   the screen-recording-layouts skill's nested panel). Never reshape the\n   whole scene to guess past it.\n\n**Before a film's cut is reviewed.** From this skill's folder, run\n`python3 scripts/cut_check.py <file>` on `{ scene, allowed?, shots?, places? }`: the cut's scene,\nits allowed spans (the screen clips placed in it, as film-cutting says), and from the project record\nthe shot each clip shows (keyed by clip id, with its size and angle when the shot list has them) and\nthe place each shows. From the scene alone it finds the joins, moves and overlays a viewer reads\nas a glitch: two takes of one shot, or two shots of one size and angle, cut together\n(`same-scene-cut`), or a cut hidden under a quick move (`hidden-take-change`); different shots of\none place are coverage and pass, a short dissolve between takes (`ghost`), a move of the picture, or of a layer covering 40% of the frame or more, that starts or stops in\none frame (`pop-move`), an overlay gone for a frame or two (`blink`), a zoom over a still picture\n(`zoom-idle`), and a screen held still inside its span (`screen-still`). Nothing inside a span is\nflagged except `screen-still`. It prints `{ ok, errors, warnings, unchecked, skipped }`: fix each error as\nfilm-cutting says and run it again, say each warning in the review, and check what it lists as\n`unchecked` on the frames of step 5, never calling it checked. On an animatic of held stills, add\n`--slides`: its stills hold on purpose, so the motion rules move to `skipped` and run on the cut,\nand the joins, overlays and screens are still checked.\n\n**Before you hand over an export.** An export can differ from its scene: another size, a dropped\nframe rate, a missing sound track, a silent last second. Probe the file with\n`picsart_media_probe`, which is free, and from this skill's folder run\n`python3 scripts/qc.py <file>` on `{ scene, probe, audio? }`: the scene you exported, the probe as\nit returned, and in `audio` only what a tool measured (loudness, true peak, the sound's length, an\nenergy curve). It prints `{ ok, errors, warnings, unchecked }`:\n\n- an `error` (the size, the fps or the length is not the scene's, the sound is missing, drifts or\n  peaks above -0.5 dBTP): make its `fix` and export again, once. Still there: deliver, and name it;\n- a `warning` (loudness away from -14 LUFS, a silent tail over 1.2 s): say it in the delivery\n  line;\n- `unchecked`: what no tool measured. Never call it checked.\n\nFrames show framing, not motion: the user judges timing and easing in REVIEW.\n\nHow many rounds you get, and what to do when they run out, is in picsart-video-rules (Evaluation). Never\nexport to \"show\" a draft that the user can review live; picsart-video-rules (Preview and export) says when\nto export.\n\n## The script, before the build\n\nBefore a short is built, voiced or generated, review its script: is it worth making, what is its\nwhy, and which lines to rewrite. A video whose footage never shows its result is dropped before it\ncosts anything. The review runs `scripts/script_review.py`; the reference below has its steps.\n\n| Read this | Before you |\n|---|---|\n| `references/script-review.md` | review a script, or report its verdict |\n| `references/readiness-verdict.md` | declare a generated video's decisive evidence before paying for it, or say whether a candidate is ready |\n"
}

SHA-256 of public snapshot: a89ecbe2dba1193e9b10ccb3ceaeebf4cd4b824ab1922f2f646366bc0a0dc751