Senior Computer Vision
Ship production-grade computer vision pipelines—from object detection tuning to distributed inference—with senior-level architecture guardrails in the agent.
Overview
Senior Computer Vision is an agent skill most often used in Build (also Ship, Operate) that guides production-ready computer vision architecture, optimization, and ML operations for inference at scale.
Install
npx skills add https://github.com/davila7/claude-code-templates --skill senior-computer-visionWhat is this skill?
- Production-first framing: scalability, 99.9% uptime targets, observability, and maintainability baked into design choice
- Advanced patterns for distributed processing, real-time low-latency paths, and ML-at-scale with monitoring
- Object detection optimization guidance for throughput, batching, and resource-aware serving
- Security and privacy defaults: input validation, encryption, access control, and audit logging
- Engineering discipline: profiling before optimize, strategic caching, retries, circuit breakers, and health checks
- Targets 10x load headroom in production-first design
- 99.9% uptime framing for reliability planning
Adoption & trust: 810 installs on skills.sh; 27.8k GitHub stars; 3/3 security scanners passed (skills.sh audits); trending (+100% hot-view momentum).
What problem does it solve?
You can train a detector in a notebook but lack a clear plan for scalable, observable, secure CV services in production.
Who is it for?
Indie builders or small teams adding detection, tracking, or embedding APIs who need senior CV production patterns without hiring a full ML platform team.
Skip if: Pure research ideation with no serving path, or teams that only need a one-off Kaggle notebook with no deployment intent.
When should I use this skill?
Designing or hardening computer vision systems, object detection serving, or ML pipelines where production scale, latency, and observability matter.
What do I get? / Deliverables
You leave with architecture patterns, optimization priorities, and reliability practices you can implement in code and infra—and revisit when latency or cost drift in production.
- Architecture recommendations for distributed, real-time, or batch CV paths
- Optimization and reliability checklist aligned to the chosen serving model
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
CV systems are implemented as backend/ML services; canonical shelf is Build because that's where models, pipelines, and inference APIs are designed. Backend subphase fits training/serving stacks, batch jobs, and real-time inference APIs rather than UI or GTM work.
Where it fits
Sketch a FastAPI or worker layout for batched object detection with clear SLAs before writing implementation tasks.
Plan how vision outputs feed search, alerting, or agent tools with stable schemas and rate limits.
Profile inference paths and decide caching, batch sizes, and hardware targets before launch.
Align image ingestion with validation, encryption, and access policies before exposing a public endpoint.
Define drift, error-rate, and latency dashboards plus retry/circuit-breaker behavior for live CV traffic.
How it compares
Architecture and MLOps judgment for vision workloads—not a turnkey training script or a generic frontend skill.
Common Questions / FAQ
Who is senior-computer-vision for?
Solo and indie builders shipping CV-backed features (SaaS APIs, agents with vision, internal tools) who want senior-level production and optimization guidance inside their coding agent.
When should I use senior-computer-vision?
During Build when designing inference APIs and data pipelines; during Ship when hardening performance, security, and test coverage; during Operate when tuning batching, monitoring drift, and cost under real traffic.
Is senior-computer-vision safe to install?
Treat it as procedural guidance only—it does not execute models by itself. Review the Security Audits panel on this Prism page and your own dependency choices before running training or inference code the agent generates.
SKILL.md
READMESKILL.md - Senior Computer Vision
# Computer Vision Architectures ## Overview World-class computer vision architectures for senior computer vision engineer. ## Core Principles ### Production-First Design Always design with production in mind: - Scalability: Handle 10x current load - Reliability: 99.9% uptime target - Maintainability: Clear, documented code - Observability: Monitor everything ### Performance by Design Optimize from the start: - Efficient algorithms - Resource awareness - Strategic caching - Batch processing ### Security & Privacy Build security in: - Input validation - Data encryption - Access control - Audit logging ## Advanced Patterns ### Pattern 1: Distributed Processing Enterprise-scale data processing with fault tolerance. ### Pattern 2: Real-Time Systems Low-latency, high-throughput systems. ### Pattern 3: ML at Scale Production ML with monitoring and automation. ## Best Practices ### Code Quality - Comprehensive testing - Clear documentation - Code reviews - Type hints ### Performance - Profile before optimizing - Monitor continuously - Cache strategically - Batch operations ### Reliability - Design for failure - Implement retries - Use circuit breakers - Monitor health ## Tools & Technologies Essential tools for this domain: - Development frameworks - Testing libraries - Deployment platforms - Monitoring solutions ## Further Reading - Research papers - Industry blogs - Conference talks - Open source projects # Object Detection Optimization ## Overview World-class object detection optimization for senior computer vision engineer. ## Core Principles ### Production-First Design Always design with production in mind: - Scalability: Handle 10x current load - Reliability: 99.9% uptime target - Maintainability: Clear, documented code - Observability: Monitor everything ### Performance by Design Optimize from the start: - Efficient algorithms - Resource awareness - Strategic caching - Batch processing ### Security & Privacy Build security in: - Input validation - Data encryption - Access control - Audit logging ## Advanced Patterns ### Pattern 1: Distributed Processing Enterprise-scale data processing with fault tolerance. ### Pattern 2: Real-Time Systems Low-latency, high-throughput systems. ### Pattern 3: ML at Scale Production ML with monitoring and automation. ## Best Practices ### Code Quality - Comprehensive testing - Clear documentation - Code reviews - Type hints ### Performance - Profile before optimizing - Monitor continuously - Cache strategically - Batch operations ### Reliability - Design for failure - Implement retries - Use circuit breakers - Monitor health ## Tools & Technologies Essential tools for this domain: - Development frameworks - Testing libraries - Deployment platforms - Monitoring solutions ## Further Reading - Research papers - Industry blogs - Conference talks - Open source projects # Production Vision Systems ## Overview World-class production vision systems for senior computer vision engineer. ## Core Principles ### Production-First Design Always design with production in mind: - Scalability: Handle 10x current load - Reliability: 99.9% uptime target - Maintainability: Clear, documented code - Observability: Monitor everything ### Performance by Design Optimize from the start: - Efficient algorithms - Resource awareness - Strategic caching - Batch processing ### Security & Privacy Build security in: - Input validation - Data encryption - Access control - Audit logging ## Advanced Patterns ### Pattern 1: Distributed Processing Enterprise-scale data processing with fault tolerance. ### Pattern 2: Real-Time Systems Low-latency, high-throughput systems. ### Pattern 3: ML at Scale Production ML with monitoring and automation. ## Best Practices ### Code Quality - Comprehensive testing - Clear documentation - Code reviews - Type hints ### Performance - Profile before optimizing - Monitor continuously - Cache strategically - Batch operations ### Reli