render-vignette

Assemble a short-form 'vignette' ad from clean product cutouts composited over a kinetic background video — birefnet cutout, then a cold-open text card + product carousel + annotated specimen-sheet end card, plus loudnorm + separate-pass music mux. FREE assembly (PIL + rsvg + FFmpeg); the recipe supplies the config and gates the paid BG/cutout/music calls to their own capabilities. Use for the vignette format.

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/ads/capabilities/render-vignette/SKILL.md

Quality

Content

72%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A dense, highly actionable runbook with exact codec/filter parameters and a clear step sequence, backed by real bundle scripts. The two weak spots are absent validation checkpoints for a failure-prone batch/mux operation and some routing-detail repetition across sections.

Suggestions

Add an explicit verification checkpoint after the mux pass (e.g., probe audio bitrate/codec with ffprobe and re-run the mux if it reads ~1 kbps) to turn the known 'garbage audio' failure mode into a validate→fix→retry loop.

Consolidate the birefnet/music routing notes so each routing decision is stated once — step 4 already covers music routing and the Gaps section restates it, which inflates token cost.

Surface the per-BG-variant batch loop as a named validation step (verify each variant's duration/resolution before mux) so the batch operation has a checkpoint rather than an implied one.

Dimension	Reasoning	Score
Conciseness	Mostly lean and technical with no concept-explanation padding, but the separate-pass-mux rationale and the birefnet/music routing story are restated across the Run, Contract, and Gaps sections and could be consolidated. Not a 3 because 'every token earns its place' is undercut by this cross-section repetition; not a 1 because nothing explains concepts Claude already knows.	2 / 3
Actionability	Gives concrete executable guidance — named scripts, exact FFmpeg params ('h264 crf20 yuv420p +faststart 30fps', 'y=(H-h)/2', loudnorm 'I=-18:TP=-2:LRA=9', '-map 0:v:0 -map 1:a:0') and numeric craft values (saturation=0.50/0.85, ~75%/65% width). Not a 2 because it is copy-paste-ready rather than pseudocode or incomplete.	3 / 3
Workflow Clarity	A clear 4-step Run sequence exists, but this is a batch operation (one filter_complex per BG variant) with a known failure mode ('silently ships 1 kbps garbage audio') and no explicit validate→fix→retry checkpoint. Per the batch/destructive guideline this caps at 2; not a 1 because the sequence itself is clearly ordered.	2 / 3
Progressive Disclosure	SKILL.md is an overview/contract runbook that points by name to four real scripts in scripts/ (strip_product_backgrounds.py, render_overlays.py, composite_variants.py, music_and_mux.py) at one level of depth, with well-organized Run/Contract/Gaps sections. Not a 2 because references are clearly signaled and appropriately split rather than inlined or nested.	3 / 3
	Total	10 / 12 Passed

Description

85%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A specific, distinctive description that names concrete assembly steps and tools and includes an explicit trigger clause. Its main weakness is narrow trigger-term coverage, relying almost entirely on the single word 'vignette'.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'birefnet cutout, then a cold-open text card + product carousel + annotated specimen-sheet end card, plus loudnorm + separate-pass music mux' with named tools (PIL + rsvg + FFmpeg) — matching the 'lists multiple specific concrete actions' anchor. Not a 2 because the action coverage is comprehensive rather than partial.	3 / 3
Completeness	Clearly states what it does ('Assemble a short-form vignette ad...') and gives an explicit trigger clause 'Use for the vignette format.', answering both what and when. Not a 2 because the 'when' is explicit rather than merely implied.	3 / 3
Trigger Term Quality	The only natural trigger term is 'vignette' ('Use for the vignette format', 'short-form vignette ad'); common variations a user might say are not covered. Not a 3 because coverage of natural terms is narrow rather than broad; not a 1 because 'vignette' and 'short-form ad' are genuinely relevant keywords.	2 / 3
Distinctiveness Conflict Risk	Highly specific niche — 'vignette' ad format with birefnet cutouts and specimen-sheet end cards — with distinct triggers unlikely to collide with other skills. Not a 2 because the triggers are clearly niche rather than overlapping with generic media skills.	3 / 3
	Total	11 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: gooseworks-ai/goose-skills
Path: skills/ads/capabilities/render-vignette/SKILL.md
Commit: 8866b2a

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.