- Home
- Skills
- Agents & Orchestration
- Applied Computer Vision
More screenshots
Works with the AI Agents you already use
Applied Computer Vision
Detect, segment, track and read images and video. 10 ordered skills, starting with Computer Vision, Object Detection, Image Segmentation.
Free
Applied Computer Vision
Example session with this skill installed
I need to build a system to detect defects on metal parts. We have 500 images of good parts but zero images of defects. Which CV approach should I use and how do I set it up?
- Read your context and instructions
- Compiled the applied computer vision
I have analyzed your request. Given you only have nominal data (good parts), the appropriate approach is Unsupervised Image Anomaly Detection.
Anomaly Detection Pipeline Design
| Component | Specification |
|---|---|
| Primary Task | Unsupervised Anomaly Detection |
| Recommended Method | Feature Memory Bank (PatchCore style) |
| Data Split | 450 images for training (nominal), 50 images + synthetic defects for validation |
| Imaging Setup | Fixed camera mount, axial lighting to minimize specular glare |
| Metrics | Image-level AUROC and Pixel-level Localization (Heatmaps) |
Implementation Details
Stabilize Imaging: Since you have no defect samples, any variation in lighting or part position will be flagged as an anomaly. Use mechanical jigs to ensure alignment.
2.
Feature Extraction: Use a pretrained backbone (e.g., WideResNet-50) to extract patch features from your 450 nominal images.
3.
Memory Bank Construction: Apply sub-sampling to the patch features to create a representative memory bank of "normal" textures.
4.
Threshold Policy: Since you lack real defects, use synthetic anomalies (e.g., CutPaste method) on 50 nominal images to set an initial threshold.
Next steps
- Verify that the metal parts are consistently oriented in all 500 images.
- Generate synthetic defects (scratches, pits) on a subset of images to test the sensitivity of the PatchCore memory bank.
- Determine the minimum defect size in pixels to choose the appropriate patch size for the feature extractor.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Vision and video developers often need a repeatable process, not a one-off prompt. This pack targets: Detect, segment, track and read images and video.
What's inside (10 skills, suggested learning order)
- Computer Vision — Detect and classify objects in images.
- Object Detection — Locate objects in images/video.
- Image Segmentation — Segment regions and instances.
- Advanced OCR — Extract text from images and documents.
- Facial Recognition and Biometrics (Regulated Use) — Compliant biometric authentication.
- AI Video Analysis — Extract events and insights from video.
- Multimodal Vision-Language Models — Use models that understand image and text.
- Image Anomaly Detection — Identify defects.
- Object Tracking — Follow objects across video.
- Pose Estimation — Detect human poses.
Why a skill pack instead of a one-off prompt
Each skill gives the agent a defined trigger, step-by-step procedure and output format, so results are consistent across sessions and easier to review.
How to use
- Install the pack in your SKILL.md-compatible agent.
- Ask for the task in plain language (for example: "detect and classify objects in images").
- Review the output against the skill's quality checks before using it.
Known limitations
- Outputs are drafts that need human review; they are not professional, legal or financial advice.
- Tools, APIs and prices change; check current vendor documentation before relying on version-specific details.
- Quality depends on the context you give the agent (data, requirements, constraints).
- Tested only where stated in the Compatibility note.
Also by Rodrigo Santos
Specialized Agents and Distribution; AI Agents: Fundamentals and Protocols; Model Fine-Tuning and Alignment
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Focuses on pipeline design and configuration over providing weights. Biometric skills are engineering gates, not legal counsel. VLM skills limited by provider downscaling and compression.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
- Free with an Agensi account
Nothing to install. It runs in your AI agent.
Instruction-only skills. No network access, shell commands or file writes are required; each skill reads the context you provide and returns a document.