Simplify the transformation skill#7
Conversation
njb90
left a comment
njb90
left a comment
There was a problem hiding this comment.
One comment but otherwise it is definitely an improvement.
strausr
left a comment
strausr
left a comment
There was a problem hiding this comment.
I ran 15 prompts against the new skill (resize, face crop, bg removal, gen fill / 16:9, w_auto, named transforms, baseline strategy, video trim + mute, debugging, g_auto)
First, using only SKILL.md + references, then checking whether the docs-first path (llms.txt) fills the gaps.
Result: path works well (resize, fill crop, overlays, bg removal, f_auto:video, debugging). Without a doc fetch, several common cases were partial or failed: face-focused crops (no g_face), gen fill, w_auto + Client Hints setup, named/baseline URL shape, video trim, and invalid g_auto + c_scale.
Left some comments on where I found some issues.
# Conflicts: # skills/cloudinary-transformations/SKILL.md # skills/cloudinary-transformations/references/advanced-features.md # skills/cloudinary-transformations/references/debugging.md # skills/cloudinary-transformations/references/examples.md # skills/cloudinary-transformations/references/named-transformations.md # skills/cloudinary-transformations/references/responsive-images.md # skills/cloudinary-transformations/references/transformation-costs.md
# Conflicts: # skills/cloudinary-transformations/SKILL.md # skills/cloudinary-transformations/references/advanced-features.md
sharon-yelenik
left a comment
sharon-yelenik
left a comment
There was a problem hiding this comment.
Looks good. Some minor thoughts
|
|
||
| ## Default Decisions | ||
|
|
||
| - Add `f_auto/q_auto` near the end for most production image URLs. |
There was a problem hiding this comment.
Maybe we should point out that the / syntax is required, if we want to ensure that we don't fall into comma-separated.
| For resize/crop requests, determine: | ||
| - At least one dimension. | ||
| - Whether the result must fill a fixed box, fit inside a box, pad empty space, or only limit maximum size. | ||
| - Whether the crop should favor faces, a known subject, center, a compass position, or Cloudinary automatic gravity. |
There was a problem hiding this comment.
"a known subject" isn't that add-on specific?
|
|
||
| Choose focal point: | ||
| - Use automatic gravity for varied content. | ||
| - Use face/person/object-specific gravity only when it matches the asset and task. |
There was a problem hiding this comment.
same comment - isn't object-specific gravity available only via an add-on? What's the best way to handle that?
|
|
||
| - Bare dimensions can imply default behavior. Make the crop mode explicit. | ||
| - `c_scale` with both width and height can distort if aspect ratios differ. | ||
| - Qualifiers belong in the same component as the action they modify. |
There was a problem hiding this comment.
Did we state that components are separated by / and qualifiers within components by , ?
There was a problem hiding this comment.
See bullet below where we reference comma-separated components. This might not be consistent with the way we state our conventions.
5 participants
Simplified cloudinary-transformations into a compact agent playbook that treats Cloudinary docs as the source of truth for exact syntax, current limitations, and costs.
Key changes:
- agent-playbook.md
- debugging-playbook.md
- cost-and-caching.md