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.
Install with Tessl CLI
npx tessl i github:alirezarezvani/claude-skills --skill senior-computer-visionOverall
score
90%
Does it follow best practices?
Validation for skill structure
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that comprehensively covers the computer vision domain with specific architectures, frameworks, and use cases. It uses proper third-person voice, includes explicit trigger guidance, and provides enough technical detail to distinguish it from general machine learning or other AI-related skills. The description balances technical depth with natural language triggers effectively.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and technologies: 'object detection, image segmentation, visual AI systems', specific architectures (CNN, Vision Transformer, YOLO, Faster R-CNN, DETR, Mask R-CNN, SAM), and deployment tools (ONNX, TensorRT). Also names specific frameworks (PyTorch, torchvision, Ultralytics, Detectron2, MMDetection). | 3 / 3 |
Completeness | Clearly answers both what (computer vision engineering for detection, segmentation, specific architectures and frameworks) AND when with explicit 'Use when...' clause covering four trigger scenarios: 'building detection pipelines, training custom models, optimizing inference, or deploying vision systems'. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'computer vision', 'object detection', 'image segmentation', 'YOLO', 'detection pipelines', 'training custom models', 'inference', 'vision systems'. Includes both high-level concepts and specific framework names users would mention. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche focused specifically on computer vision and object detection. The specific model architectures (YOLO, Faster R-CNN, DETR, SAM) and frameworks (Detectron2, MMDetection, Ultralytics) create clear boundaries that wouldn't conflict with general ML or other AI skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, well-organized computer vision skill with excellent actionability and workflow clarity. The main weakness is some verbosity in sections like 'Core Expertise' and 'Tech Stack' that explain concepts Claude already knows. The progressive disclosure is well-executed with clear navigation and appropriate external references.
Suggestions
Remove or significantly condense the 'Core Expertise' section - Claude already knows what object detection and segmentation are
Consider condensing the 'Tech Stack' table into a more compact format or moving it to a reference file, as framework names alone don't add actionable value
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is reasonably efficient but includes some unnecessary content like the 'Core Expertise' section listing capabilities Claude already knows, and verbose tables that could be condensed. The tech stack table and some explanatory text add tokens without proportional value. | 2 / 3 |
Actionability | Provides fully executable commands throughout with copy-paste ready bash commands, Python snippets, and YAML configurations. The common commands section and workflow steps contain concrete, specific guidance that can be immediately used. | 3 / 3 |
Workflow Clarity | All three workflows have clear numbered steps with explicit validation checkpoints (e.g., 'Verify ONNX model', 'Validate on test set', dataset analysis reports). The optimization workflow includes benchmark comparisons and the dataset workflow includes cleaning/validation steps before proceeding. | 3 / 3 |
Progressive Disclosure | Well-structured with a clear table of contents, concise main content, and appropriate references to external files (references/computer_vision_architectures.md, etc.) for detailed information. References are one level deep and clearly signaled with descriptive bullet points. | 3 / 3 |
Total | 11 / 12 Passed |
Validation
75%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 12 / 16 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (532 lines); consider splitting into references/ and linking | Warning |
metadata_version | 'metadata' field is not a dictionary | Warning |
license_field | 'license' field is missing | Warning |
body_steps | No step-by-step structure detected (no ordered list); consider adding a simple workflow | Warning |
Total | 12 / 16 Passed | |
- Reviewed
Table of Contents
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.