Name: senior-computer-vision
Rating: 64 (1 reviews)
Author: alirezarezvani

senior-computer-vision

Computer vision engineering skill for object detection, image segmentation, and visual AI systems. Covers CNN and Vision Transformer architectures, YOLO/Faster R-CNN/DETR detection, Mask R-CNN/SAM segmentation, and production deployment with ONNX/TensorRT. Includes PyTorch, torchvision, Ultralytics, Detectron2, and MMDetection frameworks. Use when building detection pipelines, training custom models, optimizing inference, or deploying vision systems.

Quality

76%

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 ./.gemini/skills/senior-computer-vision/SKILL.md

The canonical home for this skill is senior-computer-vision in alirezarezvani/claude-skills

Quality

Content

65%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, well-structured CV skill with strong executable guidance across three workflows, undermined by missing validation on destructive dataset operations and bundle references (scripts/, references/) that point to non-existent files.

Suggestions

Add a validation/checkpoint step to Workflow 3 (e.g., verify image counts before and after cleaning, dry-run augmentations on a sample, confirm split integrity) so destructive batch operations have a feedback loop.

Either provide the referenced bundle files (scripts/vision_model_trainer.py, scripts/inference_optimizer.py, scripts/dataset_pipeline_builder.py, references/*.md) or remove the references and inline only what is actually usable.

Move the inlined reference catalogs (Architecture Selection Guide, Performance Targets, augmentation config) into the references/*.md files so SKILL.md stays a lean overview with clearly signaled one-level-deep pointers.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient—commands, tables, and config snippets rather than prose explaining concepts Claude knows—though the Core Expertise, Tech Stack, and Architecture Selection Guide sections overlap somewhat and could be tightened.	4 / 5
Actionability	It provides concrete, copy-pasteable commands throughout (Quick Start, per-step bash, inline python one-liners like onnx.checker and coremltools), with the caveat that the hero scripts referenced (scripts/vision_model_trainer.py, inference_optimizer.py, dataset_pipeline_builder.py) are not actually bundled.	4 / 5
Workflow Clarity	Workflows are well-sequenced with six numbered steps each and some verification (onnx.checker, val/evaluate), but Workflow 3 performs destructive/batch operations—removing corrupted/duplicate images, augmenting, splitting—without any validation checkpoint or feedback loop, which caps workflow_clarity at 3 per the rubric.	3 / 5
Progressive Disclosure	Structure is decent (TOC, sectioned workflows, a signaled Reference Documentation link and Resources list), but substantial reference material (architecture tables, augmentation YAML, performance targets) is inlined rather than moved to the referenced files, and none of the referenced references/.md or scripts/ bundle files actually exist, so navigation does not resolve.	3 / 5
	Total	14 / 20 Passed

Description

87%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, specific description that clearly states capabilities and provides explicit "Use when" triggers in third person. Minor room to lead with action verbs over framework/model catalogs.

Dimension	Reasoning	Score
Specificity	Names concrete capabilities ("object detection, image segmentation... production deployment with ONNX/TensorRT") and lists specific architectures, but leans slightly toward cataloging frameworks/models rather than enumerating many distinct action verbs, so it sits just below the comprehensive anchor 5.	4 / 5
Completeness	It explicitly answers both what ("Computer vision engineering skill for object detection, image segmentation...") and when with concrete trigger phrases ("Use when building detection pipelines, training custom models, optimizing inference, or deploying vision systems"), matching the anchor 5 example.	5 / 5
Trigger Term Quality	Natural terms a user would say are present ("object detection", "image segmentation", "building detection pipelines", "training custom models", "optimizing inference", "deploying vision systems"), giving good keyword coverage, though a few common synonyms/variations are missing.	4 / 5
Distinctiveness Conflict Risk	It carves a clear CV-engineering niche with distinct triggers (detection pipelines, inference optimization, vision deployment) and specific stack mentions, giving minimal conflict risk with unrelated skills.	5 / 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
referenced_paths_exist	Referenced path issues: 21 missing	Warning

	Total	15 / 16 Passed

Repository: alirezarezvani/claude-skills
Path: .gemini/skills/senior-computer-vision/SKILL.md
Commit: 19392f7

Reviewed: ~14 hours 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.