Update to Picsart
Snapshot Nov 7, 2026 · 00:09 UTC · version 2.4.0
Collection source: downloaded plugin package.
Instructions updated for spoken-captions
Skill instructions — Now says: rewrite the word times.** A caption carries only what is said: never a speaker's name the viewer has not heard, a stage direction or a row...
Observed in instructions or declared skills. Runtime behavior has not been tested.
Up to 3 plugins free · no card required.
All fields and original values.
Skill instructions
Oct 27, 2026 · 06:06 UTC
rewrite the word times.** `transcripts` maps each spoken source's asset id to its saved transcript. Ask for `captions`,
Nov 7, 2026 · 00:09 UTC
rewrite the word times.** A caption carries only what is said: never a speaker's name the viewer has not heard, a stage direction or a row label. `transcripts` maps each spoken source's asset id to its transcript (its file, where you run...
Supporting files
Oct 27, 2026 · 06:06 UTC
[{"relative_path":"scripts/reel_text.py","size_in_bytes":15686}]
Nov 7, 2026 · 00:09 UTC
[{"relative_path":"scripts/reel_text.py","size_in_bytes":24584}]
Compare saved observations
Download comparison JSONSKILL.md text changes
− removed · + added
--- before +++ after @@ -6,7 +6,8 @@ # Captions Every word on screen comes from a transcript, with its times as the transcript gives them. **Never -rewrite the word times.** +rewrite the word times.** A caption carries only what is said: never a speaker's name the viewer has +not heard, a stage direction or a row label. ## The transcript @@ -63,7 +64,9 @@ } ``` -`transcripts` maps each spoken source's asset id to its saved transcript. Ask for `captions`, +`transcripts` maps each spoken source's asset id to its transcript (its file, where you run +commands). The script builds the words into the scene: where you cannot run commands, stop as +picsart-video-rules says for a script that builds the scene. Ask for `captions`, `highlights` or both. The script reads each clip's place on the reel and its trim from the scene, so a clip the user retrimmed in the editor is read as they left it, and it prints the scene with the new layers on top. A phrase it had to cut short comes back in `notes`. @@ -78,6 +81,34 @@ renderer fits to the measured words. Pass `highlightStyle` to change the look; the font follows the scene-building-basics skill. +### Word cards and the hook (a reference's format) + +When the record's `format` sets the captions as word cards, or a hook card, ask the script for +them instead of `captions`: + +```json +{ + "scene": "scene.json", + "transcripts": {"talk": "transcript-talk.json"}, + "cards": {"words": 3, "seam": 960, "style": {"font": {"family": "bebas-neue-regular", "size": 100}}}, + "hook": {"lines": ["HE'S PARALYZED.", "AND HE WON."], "box": "#C8161D", "size": 96, + "seam": 960, "gap": 70, "length": 3.2} +} +``` + +- **Cards**: one `captions` layer of 1 to 3 words each (`words`), a card closing early on a word + ending in `. , ? !`. The words keep their spoken times; each card holds toward the next, so they + never blink. `seam` puts the cards on the seam of a split. On a split, the words come from the + track with the sound (the top one); the muted bottom track is never read. A segment without + word times is one card over its own span, and the last card stops where the speech does, never + over the end card. +- **Hook**: a colour box that hugs the lines (`hook-box`) and the lines on it (`hook-text`), from + the reel's start for `length` seconds, the box's bottom `gap` px above the seam. The lines are + the ones the map approved, never written here. More than three lines, or a line over 22 + characters, comes back in `notes`. + +The style, the box colour, the size and the length come from the format, never from taste. + ### Choosing highlights Highlights are a few lines worth reading, not subtitles.
Full technical diff · 2 changed fields
changed /included_files
[
{
"relative_path": "scripts/reel_text.py",
"size_in_bytes": 15686
}
][
{
"relative_path": "scripts/reel_text.py",
"size_in_bytes": 24584
}
]changed /skill_md_contents
"---\nname: spoken-captions\ndescription: Put spoken words on screen from the transcript - word-timed captions on a generated voice-over or on the user's own talking footage, and on a cut reel, captions per clip and a few highlight phrases timed to the words they call out, placed with scripts/reel_text.py. Use when a video is watched muted and its speech must be read, or when a short's key lines should land on screen.\n---\n\n# Captions\n\nEvery word on screen comes from a transcript, with its times as the transcript gives them. **Never\nrewrite the word times.**\n\n## The transcript\n\n1. **Probe before you transcribe.** Check `hasAudio` from `picsart_media_probe` first, and\n skip only when it reports `hasAudio: false`, as the footage-analysis skill says.\n2. Transcribe with `picsart_media_transcribe`:\n - a voice-over you generated: **each generated voice line**. It is our own clean audio, so the\n word edges are exact.\n - the user's own footage: each source once, as the footage-analysis skill says. Reuse that\n transcript; never transcribe a cut.\n\nThe tool's `notes` name any segment it returned without word times. Caption that segment as one\nblock over its own start and end; it cannot carry a highlight.\n\n## Captions on a voice-over\n\nPlace each line as a `captions` layer through `picsart_scene_patch`:\n`content: { kind: \"captions\", transcript, timeOffset }`, where `transcript` is the tool's\n`transcript` as returned (it is already shaped for a captions layer) and `timeOffset` is the\nline's start in the video.\n\n2–4 words per caption. The caption style's `offset` is measured from the frame center\n(`[0, -670]` puts it at y ≈ 290 in a 1080×1920 frame, below the top 250 px the platforms cover); keep captions in a dark band of the\nlayout, never over a panel or the focus point.\n\n**Readable over any picture, and the spoken word lit.** A caption crosses moving footage, so give\nits style an outline and light the word being spoken:\n\n```json\n\"style\": {\n \"effects\": [{\"id\": \"stroke\", \"params\": {\"width\": \"<px>\", \"color\": \"<outline colour>\"}}],\n \"decorations\": [{\"target\": \"highlighted\", \"place\": \"text\", \"color\": \"<highlight colour>\"}]\n}\n```\n\nRead the effect's id and params from `picsart_scene_capabilities` (`effects`) rather than\nfrom here; this shape passed `picsart_scene_validate` on 2026-10-05. One highlight colour\nfor the whole video, the brand's when it has one.\n\n## Words on a cut reel\n\nA reel cut from the user's footage plays each source in pieces, so its words are placed from the\nscene the user approved, with `python3 scripts/reel_text.py words.json`:\n\n```json\n{\n \"scene\": \"scene.json\",\n \"transcripts\": {\"talk\": \"transcript-talk.json\"},\n \"captions\": {\"style\": {\"font\": {\"size\": 64}, \"offset\": [0, 440]}},\n \"highlights\": [\n {\"clip\": 0, \"say\": \"revenue tripled\", \"text\": \"Revenue tripled\"},\n {\"clip\": 4, \"say\": \"nobody tells you this\"}\n ]\n}\n```\n\n`transcripts` maps each spoken source's asset id to its saved transcript. Ask for `captions`,\n`highlights` or both. The script reads each clip's place on the reel and its trim from the scene,\nso a clip the user retrimmed in the editor is read as they left it, and it prints the scene with\nthe new layers on top. A phrase it had to cut short comes back in `notes`.\n\n- **Captions** become one `captions` layer per spoken clip, open while that clip plays its\n source, with the source's transcript as returned. A clip held past its `trim.out` freezes on its\n last frame, so its captions close where the trim ends. The reel ends at the composition's\n `duration`, or with none at the track layer's: captions stop there. A photo, a `muted` clip, a\n shot without a transcript or a clip past the reel's end gets none, and a highlight on a muted\n clip or one past the end is refused.\n- **Highlights** all go on one `captions` layer, one segment per phrase, each on a pill the\n renderer fits to the measured words. Pass `highlightStyle` to change the look; the font follows\n the scene-building-basics skill.\n\n### Choosing highlights\n\nHighlights are a few lines worth reading, not subtitles.\n\n- **2–4 per reel, 2–6 words each, verbatim from the transcript.** `say` is what was said;\n `text` may drop a filler word or fix its case, never change what it says.\n- Pick the hook, the number, the rehook (the turn near the middle, as the retention-hooks skill names\n it) and the payoff: the words a viewer should remember. Skip\n greetings, filler and anything already said on screen.\n- **Draft them before you place them:** each phrase with its clip, so the use case can offer\n them. The reel is complete without them.\n- One at a time. The script holds each from its first word to half a second past its last,\n 1.5–3 s, and never past its clip or into the next phrase.\n- **Big, high-contrast, inside the platform's safe band and off the face.** The use case names\n the band. With captions on too, keep the two in different bands.\n\nCheck the result as the quality-review skill says, with a frame at each phrase's `start`.\n""---\nname: spoken-captions\ndescription: Put spoken words on screen from the transcript - word-timed captions on a generated voice-over or on the user's own talking footage, and on a cut reel, captions per clip and a few highlight phrases timed to the words they call out, placed with scripts/reel_text.py. Use when a video is watched muted and its speech must be read, or when a short's key lines should land on screen.\n---\n\n# Captions\n\nEvery word on screen comes from a transcript, with its times as the transcript gives them. **Never\nrewrite the word times.** A caption carries only what is said: never a speaker's name the viewer has\nnot heard, a stage direction or a row label.\n\n## The transcript\n\n1. **Probe before you transcribe.** Check `hasAudio` from `picsart_media_probe` first, and\n skip only when it reports `hasAudio: false`, as the footage-analysis skill says.\n2. Transcribe with `picsart_media_transcribe`:\n - a voice-over you generated: **each generated voice line**. It is our own clean audio, so the\n word edges are exact.\n - the user's own footage: each source once, as the footage-analysis skill says. Reuse that\n transcript; never transcribe a cut.\n\nThe tool's `notes` name any segment it returned without word times. Caption that segment as one\nblock over its own start and end; it cannot carry a highlight.\n\n## Captions on a voice-over\n\nPlace each line as a `captions` layer through `picsart_scene_patch`:\n`content: { kind: \"captions\", transcript, timeOffset }`, where `transcript` is the tool's\n`transcript` as returned (it is already shaped for a captions layer) and `timeOffset` is the\nline's start in the video.\n\n2–4 words per caption. The caption style's `offset` is measured from the frame center\n(`[0, -670]` puts it at y ≈ 290 in a 1080×1920 frame, below the top 250 px the platforms cover); keep captions in a dark band of the\nlayout, never over a panel or the focus point.\n\n**Readable over any picture, and the spoken word lit.** A caption crosses moving footage, so give\nits style an outline and light the word being spoken:\n\n```json\n\"style\": {\n \"effects\": [{\"id\": \"stroke\", \"params\": {\"width\": \"<px>\", \"color\": \"<outline colour>\"}}],\n \"decorations\": [{\"target\": \"highlighted\", \"place\": \"text\", \"color\": \"<highlight colour>\"}]\n}\n```\n\nRead the effect's id and params from `picsart_scene_capabilities` (`effects`) rather than\nfrom here; this shape passed `picsart_scene_validate` on 2026-10-05. One highlight colour\nfor the whole video, the brand's when it has one.\n\n## Words on a cut reel\n\nA reel cut from the user's footage plays each source in pieces, so its words are placed from the\nscene the user approved, with `python3 scripts/reel_text.py words.json`:\n\n```json\n{\n \"scene\": \"scene.json\",\n \"transcripts\": {\"talk\": \"transcript-talk.json\"},\n \"captions\": {\"style\": {\"font\": {\"size\": 64}, \"offset\": [0, 440]}},\n \"highlights\": [\n {\"clip\": 0, \"say\": \"revenue tripled\", \"text\": \"Revenue tripled\"},\n {\"clip\": 4, \"say\": \"nobody tells you this\"}\n ]\n}\n```\n\n`transcripts` maps each spoken source's asset id to its transcript (its file, where you run\ncommands). The script builds the words into the scene: where you cannot run commands, stop as\npicsart-video-rules says for a script that builds the scene. Ask for `captions`,\n`highlights` or both. The script reads each clip's place on the reel and its trim from the scene,\nso a clip the user retrimmed in the editor is read as they left it, and it prints the scene with\nthe new layers on top. A phrase it had to cut short comes back in `notes`.\n\n- **Captions** become one `captions` layer per spoken clip, open while that clip plays its\n source, with the source's transcript as returned. A clip held past its `trim.out` freezes on its\n last frame, so its captions close where the trim ends. The reel ends at the composition's\n `duration`, or with none at the track layer's: captions stop there. A photo, a `muted` clip, a\n shot without a transcript or a clip past the reel's end gets none, and a highlight on a muted\n clip or one past the end is refused.\n- **Highlights** all go on one `captions` layer, one segment per phrase, each on a pill the\n renderer fits to the measured words. Pass `highlightStyle` to change the look; the font follows\n the scene-building-basics skill.\n\n### Word cards and the hook (a reference's format)\n\nWhen the record's `format` sets the captions as word cards, or a hook card, ask the script for\nthem instead of `captions`:\n\n```json\n{\n \"scene\": \"scene.json\",\n \"transcripts\": {\"talk\": \"transcript-talk.json\"},\n \"cards\": {\"words\": 3, \"seam\": 960, \"style\": {\"font\": {\"family\": \"bebas-neue-regular\", \"size\": 100}}},\n \"hook\": {\"lines\": [\"HE'S PARALYZED.\", \"AND HE WON.\"], \"box\": \"#C8161D\", \"size\": 96,\n \"seam\": 960, \"gap\": 70, \"length\": 3.2}\n}\n```\n\n- **Cards**: one `captions` layer of 1 to 3 words each (`words`), a card closing early on a word\n ending in `. , ? !`. The words keep their spoken times; each card holds toward the next, so they\n never blink. `seam` puts the cards on the seam of a split. On a split, the words come from the\n track with the sound (the top one); the muted bottom track is never read. A segment without\n word times is one card over its own span, and the last card stops where the speech does, never\n over the end card.\n- **Hook**: a colour box that hugs the lines (`hook-box`) and the lines on it (`hook-text`), from\n the reel's start for `length` seconds, the box's bottom `gap` px above the seam. The lines are\n the ones the map approved, never written here. More than three lines, or a line over 22\n characters, comes back in `notes`.\n\nThe style, the box colour, the size and the length come from the format, never from taste.\n\n### Choosing highlights\n\nHighlights are a few lines worth reading, not subtitles.\n\n- **2–4 per reel, 2–6 words each, verbatim from the transcript.** `say` is what was said;\n `text` may drop a filler word or fix its case, never change what it says.\n- Pick the hook, the number, the rehook (the turn near the middle, as the retention-hooks skill names\n it) and the payoff: the words a viewer should remember. Skip\n greetings, filler and anything already said on screen.\n- **Draft them before you place them:** each phrase with its clip, so the use case can offer\n them. The reel is complete without them.\n- One at a time. The script holds each from its first word to half a second past its last,\n 1.5–3 s, and never past its clip or into the next phrase.\n- **Big, high-contrast, inside the platform's safe band and off the face.** The use case names\n the band. With captions on too, keep the two in different bands.\n\nCheck the result as the quality-review skill says, with a frame at each phrase's `start`.\n"Full snapshot data
{
"description": "Put spoken words on screen from the transcript - word-timed captions on a generated voice-over or on the user's own talking footage, and on a cut reel, captions per clip and a few highlight phrases timed to the words they call out, placed with scripts/reel_text.py. Use when a video is watched muted and its speech must be read, or when a short's key lines should land on screen.",
"included_files": [
{
"relative_path": "scripts/reel_text.py",
"size_in_bytes": 24584
}
],
"name": "spoken-captions",
"skill_md_contents": "---\nname: spoken-captions\ndescription: Put spoken words on screen from the transcript - word-timed captions on a generated voice-over or on the user's own talking footage, and on a cut reel, captions per clip and a few highlight phrases timed to the words they call out, placed with scripts/reel_text.py. Use when a video is watched muted and its speech must be read, or when a short's key lines should land on screen.\n---\n\n# Captions\n\nEvery word on screen comes from a transcript, with its times as the transcript gives them. **Never\nrewrite the word times.** A caption carries only what is said: never a speaker's name the viewer has\nnot heard, a stage direction or a row label.\n\n## The transcript\n\n1. **Probe before you transcribe.** Check `hasAudio` from `picsart_media_probe` first, and\n skip only when it reports `hasAudio: false`, as the footage-analysis skill says.\n2. Transcribe with `picsart_media_transcribe`:\n - a voice-over you generated: **each generated voice line**. It is our own clean audio, so the\n word edges are exact.\n - the user's own footage: each source once, as the footage-analysis skill says. Reuse that\n transcript; never transcribe a cut.\n\nThe tool's `notes` name any segment it returned without word times. Caption that segment as one\nblock over its own start and end; it cannot carry a highlight.\n\n## Captions on a voice-over\n\nPlace each line as a `captions` layer through `picsart_scene_patch`:\n`content: { kind: \"captions\", transcript, timeOffset }`, where `transcript` is the tool's\n`transcript` as returned (it is already shaped for a captions layer) and `timeOffset` is the\nline's start in the video.\n\n2–4 words per caption. The caption style's `offset` is measured from the frame center\n(`[0, -670]` puts it at y ≈ 290 in a 1080×1920 frame, below the top 250 px the platforms cover); keep captions in a dark band of the\nlayout, never over a panel or the focus point.\n\n**Readable over any picture, and the spoken word lit.** A caption crosses moving footage, so give\nits style an outline and light the word being spoken:\n\n```json\n\"style\": {\n \"effects\": [{\"id\": \"stroke\", \"params\": {\"width\": \"<px>\", \"color\": \"<outline colour>\"}}],\n \"decorations\": [{\"target\": \"highlighted\", \"place\": \"text\", \"color\": \"<highlight colour>\"}]\n}\n```\n\nRead the effect's id and params from `picsart_scene_capabilities` (`effects`) rather than\nfrom here; this shape passed `picsart_scene_validate` on 2026-10-05. One highlight colour\nfor the whole video, the brand's when it has one.\n\n## Words on a cut reel\n\nA reel cut from the user's footage plays each source in pieces, so its words are placed from the\nscene the user approved, with `python3 scripts/reel_text.py words.json`:\n\n```json\n{\n \"scene\": \"scene.json\",\n \"transcripts\": {\"talk\": \"transcript-talk.json\"},\n \"captions\": {\"style\": {\"font\": {\"size\": 64}, \"offset\": [0, 440]}},\n \"highlights\": [\n {\"clip\": 0, \"say\": \"revenue tripled\", \"text\": \"Revenue tripled\"},\n {\"clip\": 4, \"say\": \"nobody tells you this\"}\n ]\n}\n```\n\n`transcripts` maps each spoken source's asset id to its transcript (its file, where you run\ncommands). The script builds the words into the scene: where you cannot run commands, stop as\npicsart-video-rules says for a script that builds the scene. Ask for `captions`,\n`highlights` or both. The script reads each clip's place on the reel and its trim from the scene,\nso a clip the user retrimmed in the editor is read as they left it, and it prints the scene with\nthe new layers on top. A phrase it had to cut short comes back in `notes`.\n\n- **Captions** become one `captions` layer per spoken clip, open while that clip plays its\n source, with the source's transcript as returned. A clip held past its `trim.out` freezes on its\n last frame, so its captions close where the trim ends. The reel ends at the composition's\n `duration`, or with none at the track layer's: captions stop there. A photo, a `muted` clip, a\n shot without a transcript or a clip past the reel's end gets none, and a highlight on a muted\n clip or one past the end is refused.\n- **Highlights** all go on one `captions` layer, one segment per phrase, each on a pill the\n renderer fits to the measured words. Pass `highlightStyle` to change the look; the font follows\n the scene-building-basics skill.\n\n### Word cards and the hook (a reference's format)\n\nWhen the record's `format` sets the captions as word cards, or a hook card, ask the script for\nthem instead of `captions`:\n\n```json\n{\n \"scene\": \"scene.json\",\n \"transcripts\": {\"talk\": \"transcript-talk.json\"},\n \"cards\": {\"words\": 3, \"seam\": 960, \"style\": {\"font\": {\"family\": \"bebas-neue-regular\", \"size\": 100}}},\n \"hook\": {\"lines\": [\"HE'S PARALYZED.\", \"AND HE WON.\"], \"box\": \"#C8161D\", \"size\": 96,\n \"seam\": 960, \"gap\": 70, \"length\": 3.2}\n}\n```\n\n- **Cards**: one `captions` layer of 1 to 3 words each (`words`), a card closing early on a word\n ending in `. , ? !`. The words keep their spoken times; each card holds toward the next, so they\n never blink. `seam` puts the cards on the seam of a split. On a split, the words come from the\n track with the sound (the top one); the muted bottom track is never read. A segment without\n word times is one card over its own span, and the last card stops where the speech does, never\n over the end card.\n- **Hook**: a colour box that hugs the lines (`hook-box`) and the lines on it (`hook-text`), from\n the reel's start for `length` seconds, the box's bottom `gap` px above the seam. The lines are\n the ones the map approved, never written here. More than three lines, or a line over 22\n characters, comes back in `notes`.\n\nThe style, the box colour, the size and the length come from the format, never from taste.\n\n### Choosing highlights\n\nHighlights are a few lines worth reading, not subtitles.\n\n- **2–4 per reel, 2–6 words each, verbatim from the transcript.** `say` is what was said;\n `text` may drop a filler word or fix its case, never change what it says.\n- Pick the hook, the number, the rehook (the turn near the middle, as the retention-hooks skill names\n it) and the payoff: the words a viewer should remember. Skip\n greetings, filler and anything already said on screen.\n- **Draft them before you place them:** each phrase with its clip, so the use case can offer\n them. The reel is complete without them.\n- One at a time. The script holds each from its first word to half a second past its last,\n 1.5–3 s, and never past its clip or into the next phrase.\n- **Big, high-contrast, inside the platform's safe band and off the face.** The use case names\n the band. With captions on too, keep the two in different bands.\n\nCheck the result as the quality-review skill says, with a frame at each phrase's `start`.\n"
}SHA-256 of public snapshot: bcb24b62353f65f9c401e940ba7c7f9d2f5e0a18adcbee309928d4ef8f4efcf3