gemini-3-multimodal

Process multimodal inputs (images, video, audio, PDFs) with Gemini 3 Pro. Covers image understanding, video analysis, audio processing, document extraction, media resolution control, OCR, and token optimization. Use when analyzing images, processing video, transcribing audio, extracting PDF content, or working with multimodal data.

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/gemini-3-multimodal/SKILL.md

Quality

Content

68%Weight 40%Scale 1-5

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

A thorough, highly actionable multimodal skill with executable Python/Node examples and well-signaled one-level references to real bundle files. Its weaknesses are verbosity (redundant/padded sections and full per-task dual-language duplication in SKILL.md) and missing validation/feedback loops around its batch and destructive file operations.

Suggestions

Trim padded sections ("Key Capabilities", checkmark "Summary", "Ready to analyze...?") and move full per-task Node.js examples into the reference files, keeping one canonical example per task in SKILL.md.

Add validation/feedback loops to the batch PDF pipeline (per-file try/except, error logging, result verification) and a confirmation/guard step before the destructive "Delete all files" loop so workflow clarity can exceed 3.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient executable guidance but carries padded/redundant sections—"Key Capabilities" duplicating the task list, a checkmark "Summary", "Common Use Cases" re-deriving code already shown, and every task duplicated in both Python and Node.js—so it could be tightened considerably.	3 / 5
Actionability	Every task ships copy-paste-ready, fully executable Python and Node.js code covering the common cases, plus concrete resolution/cost tables and token-usage inspection, matching the anchor for fully executable guidance.	5 / 5
Workflow Clarity	Each task sequences upload → wait-for-processing → FAILED check → analyze, but the batch "Process all PDFs in directory" and destructive "Delete all files" operations lack per-item validation, error recovery, or confirmation, which caps workflow clarity at 3 per the batch/destructive guideline.	3 / 5
Progressive Disclosure	References and scripts are real, one level deep, and clearly signaled with "See: references/...md" plus a consolidated References section; the only gap is that substantial dual-language task examples are inlined in SKILL.md rather than pushed into the already-present reference files.	4 / 5
	Total	15 / 20 Passed

Description

88%Weight 40%Scale 1-5

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 strong, well-structured description that answers both what and when with concrete, mostly natural trigger phrases and a clear Gemini-3-Pro multimodal-input niche. It could be sharpened by adding file extensions and synonyms for trigger terms and tightening the generic "multimodal data" phrasing.

Suggestions

Add file extensions and synonyms to the trigger clause (e.g. ".jpg/.png images, .mp4 video, .mp3/.wav audio, .pdf files") to reach comprehensive trigger-term coverage.

Replace the generic "working with multimodal data" with more specific triggers to reduce overlap risk with sibling gemini-3 skills.

Dimension	Reasoning	Score
Specificity	The description enumerates multiple concrete capabilities—"image understanding, video analysis, audio processing, document extraction, media resolution control, OCR, and token optimization"—and ties them to specific actions in the trigger clause, giving comprehensive coverage.	5 / 5
Completeness	It explicitly states both what ("Process multimodal inputs ... with Gemini 3 Pro. Covers ...") and when ("Use when analyzing images, processing video, transcribing audio, extracting PDF content, or working with multimodal data.") with concrete trigger phrases.	5 / 5
Trigger Term Quality	"analyzing images, processing video, transcribing audio, extracting PDF content, or working with multimodal data" are natural user phrases, but the description omits file extensions (e.g. .pdf, .mp4, .jpg) and synonyms that would push it to a 5.	4 / 5
Distinctiveness Conflict Risk	The "Gemini 3 Pro" multimodal-input niche is clearly scoped and distinguished from output generation, but the broad "working with multimodal data" trigger and adjacency to other gemini-3 skills leave minor overlap risk rather than minimal.	4 / 5
	Total	18 / 20 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
skill_md_line_count	SKILL.md is long (869 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: fernandezbaptiste/Skrillz
Path: skills/gemini-3-multimodal/SKILL.md
Commit: 93ed392

Reviewed: 2 days 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.