← falCONTENT HISTORY

Update to fal

Snapshot Nov 14, 2026 · 00:10 UTC · version 1.3.2

Collection source: downloaded plugin package. These snapshots do not have a confirmed matching collection source. Differences in file lists alone do not establish changes to the package.

WHAT CHANGED

Instructions updated for commercial

Skill instructions — Now says: Bundled guidance is fallback craft. For inputs, recommend_model's `input_contract` and live validation at submission are authoritative. See [the shared...

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

Save fal and check future changes in Following.

Up to 3 plugins free · no card required.

All fields and original values.

Skill instructions

Before

Sep 30, 2026 · 22:46 UTC

3. Build an approved still before video unless the user explicitly wants text-to-video exploration. ## Model Choices - Text-heavy ads/posters/UI: `openai/gpt-image-2`. - Premium product stills: `openai/gpt-image-2`, then `fal-ai/nan...

After

Nov 14, 2026 · 00:10 UTC

Bundled guidance is fallback craft. For inputs, recommend_model's `input_contract` and live validation at submission are authoritative. See [the shared production contract](../shared/production-contract.md). 3. Build a hero still, then a...

Compare saved observations

Download comparison JSON

SKILL.md text changes

− removed · + added

--- before
+++ after
@@ -8,6 +8,10 @@
 
 # Commercial Production
 
+Bundled guidance is fallback craft. For inputs, recommend_model's
+`input_contract` and live validation at submission are authoritative. See
+[the shared production contract](../shared/production-contract.md).
+
 Use when the user wants advertising, product, brand, or e-commerce media.
 Execution runs through the fal.ai MCP tools; this skill provides production
 judgment, not another command layer.
@@ -28,8 +32,9 @@
 1. Choose the route: text-to-image, reference edit, product cleanup, or
    still-to-video reveal.
 2. Use reference/edit workflows when product fidelity matters.
-3. Build an approved still before video unless the user explicitly wants
-   text-to-video exploration.
+3. Build a hero still, then animate it in the same turn (pause only if the
+   user asked to review it), unless the user explicitly wants text-to-video
+   exploration.
 4. For e-commerce batches, keep one prompt skeleton and vary one axis at a
    time: background, crop, lighting, platform, or color treatment.
 5. For text overlays, reserve clean space unless using a text-capable model
@@ -46,15 +51,13 @@
 7. Brand tone.
 8. Guardrails: preserve logo/package, no extra text, no distorted labels.
 
-## Model Choices
+## Model selection
 
-- Text-heavy ads/posters/UI: `openai/gpt-image-2`.
-- Premium product stills: `openai/gpt-image-2`, then
-  `fal-ai/nano-banana-pro`, then `fal-ai/nano-banana-2`.
-- Product/reference edits: `fal-ai/nano-banana-pro/edit`, then
-  `openai/gpt-image-2/edit`.
-- Product reveal video: `bytedance/seedance-2.0/image-to-video`.
-- Fast draft video: Grok Imagine Video endpoints.
+Use `model-routing` with this brief and actual references, then submit the
+selected `next_call`. Keep a model the user named. Bundled family suggestions
+must not override the current task-specific recommendation. Inspect one result
+before dependent continuity shots; submit independent variations together with
+`submit_jobs`. Preserve originals through any permitted repair.
 
 ## Quality Bar
 
Full technical diff · 1 changed fields

changed /skill_md_contents

BEFORE
"---\nname: commercial\ndescription: >\n  Plan and produce commercial image or video assets with fal.ai. Use for\n  product photography, ads, e-commerce batches, product reveals, lifestyle\n  commercials, background replacement, platform crops, and brand-safe prompts.\n---\n\n# Commercial Production\n\nUse when the user wants advertising, product, brand, or e-commerce media.\nExecution runs through the fal.ai MCP tools; this skill provides production\njudgment, not another command layer.\n\nLoad `model-routing` when the model is not specified.\n\n## Inputs To Collect\n\n- Product: name, category, material, color, scale, logo/packaging rules.\n- Goal: hero shot, PDP image, ad creative, product reveal, demo, lifestyle.\n- Platform: square, vertical, landscape, banner, transparent background, print.\n- Brand: premium, playful, clinical, athletic, minimal, natural, technical.\n- Source media: packshot, logo, reference scene, prior generated asset.\n- Constraints: preserve packaging, avoid new labels, no fake readable copy.\n\n## Production Workflow\n\n1. Choose the route: text-to-image, reference edit, product cleanup, or\n   still-to-video reveal.\n2. Use reference/edit workflows when product fidelity matters.\n3. Build an approved still before video unless the user explicitly wants\n   text-to-video exploration.\n4. For e-commerce batches, keep one prompt skeleton and vary one axis at a\n   time: background, crop, lighting, platform, or color treatment.\n5. For text overlays, reserve clean space unless using a text-capable model\n   with exact user-provided copy.\n\n## Prompt Order\n\n1. Product invariant: object, material, packaging, scale, logo rules.\n2. Commercial role: hero, PDP, launch teaser, social ad, demo.\n3. Setting: surface, background, props, environment.\n4. Lighting: softbox, strip light, rim light, practicals, caustics.\n5. Camera: angle, focal-length feel, macro, depth of field, motion.\n6. Composition: safe zones, crop, negative space, platform.\n7. Brand tone.\n8. Guardrails: preserve logo/package, no extra text, no distorted labels.\n\n## Model Choices\n\n- Text-heavy ads/posters/UI: `openai/gpt-image-2`.\n- Premium product stills: `openai/gpt-image-2`, then\n  `fal-ai/nano-banana-pro`, then `fal-ai/nano-banana-2`.\n- Product/reference edits: `fal-ai/nano-banana-pro/edit`, then\n  `openai/gpt-image-2/edit`.\n- Product reveal video: `bytedance/seedance-2.0/image-to-video`.\n- Fast draft video: Grok Imagine Video endpoints.\n\n## Quality Bar\n\n- Product shape, logo, material, and color are stable.\n- Generated text is absent or exact and intentional.\n- Composition supports crop and external copy.\n- Props support the product instead of competing with it.\n- The final answer includes endpoint, request id when available, output URL,\n  prompt summary, and any defects.\n"
AFTER
"---\nname: commercial\ndescription: >\n  Plan and produce commercial image or video assets with fal.ai. Use for\n  product photography, ads, e-commerce batches, product reveals, lifestyle\n  commercials, background replacement, platform crops, and brand-safe prompts.\n---\n\n# Commercial Production\n\nBundled guidance is fallback craft. For inputs, recommend_model's\n`input_contract` and live validation at submission are authoritative. See\n[the shared production contract](../shared/production-contract.md).\n\nUse when the user wants advertising, product, brand, or e-commerce media.\nExecution runs through the fal.ai MCP tools; this skill provides production\njudgment, not another command layer.\n\nLoad `model-routing` when the model is not specified.\n\n## Inputs To Collect\n\n- Product: name, category, material, color, scale, logo/packaging rules.\n- Goal: hero shot, PDP image, ad creative, product reveal, demo, lifestyle.\n- Platform: square, vertical, landscape, banner, transparent background, print.\n- Brand: premium, playful, clinical, athletic, minimal, natural, technical.\n- Source media: packshot, logo, reference scene, prior generated asset.\n- Constraints: preserve packaging, avoid new labels, no fake readable copy.\n\n## Production Workflow\n\n1. Choose the route: text-to-image, reference edit, product cleanup, or\n   still-to-video reveal.\n2. Use reference/edit workflows when product fidelity matters.\n3. Build a hero still, then animate it in the same turn (pause only if the\n   user asked to review it), unless the user explicitly wants text-to-video\n   exploration.\n4. For e-commerce batches, keep one prompt skeleton and vary one axis at a\n   time: background, crop, lighting, platform, or color treatment.\n5. For text overlays, reserve clean space unless using a text-capable model\n   with exact user-provided copy.\n\n## Prompt Order\n\n1. Product invariant: object, material, packaging, scale, logo rules.\n2. Commercial role: hero, PDP, launch teaser, social ad, demo.\n3. Setting: surface, background, props, environment.\n4. Lighting: softbox, strip light, rim light, practicals, caustics.\n5. Camera: angle, focal-length feel, macro, depth of field, motion.\n6. Composition: safe zones, crop, negative space, platform.\n7. Brand tone.\n8. Guardrails: preserve logo/package, no extra text, no distorted labels.\n\n## Model selection\n\nUse `model-routing` with this brief and actual references, then submit the\nselected `next_call`. Keep a model the user named. Bundled family suggestions\nmust not override the current task-specific recommendation. Inspect one result\nbefore dependent continuity shots; submit independent variations together with\n`submit_jobs`. Preserve originals through any permitted repair.\n\n## Quality Bar\n\n- Product shape, logo, material, and color are stable.\n- Generated text is absent or exact and intentional.\n- Composition supports crop and external copy.\n- Props support the product instead of competing with it.\n- The final answer includes endpoint, request id when available, output URL,\n  prompt summary, and any defects.\n"
Full snapshot data
{
  "description": "Plan and produce commercial image or video assets with fal.ai. Use for product photography, ads, e-commerce batches, product reveals, lifestyle commercials, background replacement, platform crops, and brand-safe prompts.\n",
  "included_files": [],
  "name": "commercial",
  "skill_md_contents": "---\nname: commercial\ndescription: >\n  Plan and produce commercial image or video assets with fal.ai. Use for\n  product photography, ads, e-commerce batches, product reveals, lifestyle\n  commercials, background replacement, platform crops, and brand-safe prompts.\n---\n\n# Commercial Production\n\nBundled guidance is fallback craft. For inputs, recommend_model's\n`input_contract` and live validation at submission are authoritative. See\n[the shared production contract](../shared/production-contract.md).\n\nUse when the user wants advertising, product, brand, or e-commerce media.\nExecution runs through the fal.ai MCP tools; this skill provides production\njudgment, not another command layer.\n\nLoad `model-routing` when the model is not specified.\n\n## Inputs To Collect\n\n- Product: name, category, material, color, scale, logo/packaging rules.\n- Goal: hero shot, PDP image, ad creative, product reveal, demo, lifestyle.\n- Platform: square, vertical, landscape, banner, transparent background, print.\n- Brand: premium, playful, clinical, athletic, minimal, natural, technical.\n- Source media: packshot, logo, reference scene, prior generated asset.\n- Constraints: preserve packaging, avoid new labels, no fake readable copy.\n\n## Production Workflow\n\n1. Choose the route: text-to-image, reference edit, product cleanup, or\n   still-to-video reveal.\n2. Use reference/edit workflows when product fidelity matters.\n3. Build a hero still, then animate it in the same turn (pause only if the\n   user asked to review it), unless the user explicitly wants text-to-video\n   exploration.\n4. For e-commerce batches, keep one prompt skeleton and vary one axis at a\n   time: background, crop, lighting, platform, or color treatment.\n5. For text overlays, reserve clean space unless using a text-capable model\n   with exact user-provided copy.\n\n## Prompt Order\n\n1. Product invariant: object, material, packaging, scale, logo rules.\n2. Commercial role: hero, PDP, launch teaser, social ad, demo.\n3. Setting: surface, background, props, environment.\n4. Lighting: softbox, strip light, rim light, practicals, caustics.\n5. Camera: angle, focal-length feel, macro, depth of field, motion.\n6. Composition: safe zones, crop, negative space, platform.\n7. Brand tone.\n8. Guardrails: preserve logo/package, no extra text, no distorted labels.\n\n## Model selection\n\nUse `model-routing` with this brief and actual references, then submit the\nselected `next_call`. Keep a model the user named. Bundled family suggestions\nmust not override the current task-specific recommendation. Inspect one result\nbefore dependent continuity shots; submit independent variations together with\n`submit_jobs`. Preserve originals through any permitted repair.\n\n## Quality Bar\n\n- Product shape, logo, material, and color are stable.\n- Generated text is absent or exact and intentional.\n- Composition supports crop and external copy.\n- Props support the product instead of competing with it.\n- The final answer includes endpoint, request id when available, output URL,\n  prompt summary, and any defects.\n"
}

SHA-256 of public snapshot: 693f1a036211ba9b15ca0abcab86778f2e1fd6102e1bb32252360d5ca36b16be