Senior Prompt Engineer
Design production-minded agentic LLM systems—prompt architecture, evaluation, and reliability patterns—for skills you ship with coding agents.
Overview
senior-prompt-engineer is an agent skill most often used in Build (also Ship security/perf, Operate monitoring) that guides production-grade agentic system and LLM evaluation design for senior prompt engineers.
Install
npx skills add https://github.com/davila7/claude-code-templates --skill senior-prompt-engineerWhat is this skill?
- Production-first framing: scalability, 99.9% uptime target, observability, maintainability
- Advanced patterns: distributed processing, real-time low-latency systems, ML at scale
- Security-by-design: validation, encryption, access control, audit logging
- Performance habits: profile-first optimization, strategic caching, batch processing
- Includes LLM evaluation frameworks section for measuring agent quality
- Documents a 99.9% uptime reliability target as a design north star
- Three advanced pattern sections: distributed processing, real-time systems, ML at scale
Adoption & trust: 764 installs on skills.sh; 27.8k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
Your agent prompts and orchestration work in demos but lack production scalability, evaluation discipline, and operational guardrails.
Who is it for?
Solo builders shipping Claude Code/Cursor agents who want a senior rubric for system design, evals, and hardening without hiring a platform team first.
Skip if: Beginners looking for copy-paste starter prompts only, or teams that need a single integration skill (billing, auth) with no agent architecture work.
When should I use this skill?
User needs senior-level agentic system design, production LLM architecture, or LLM evaluation framework guidance while building or hardening agents.
What do I get? / Deliverables
You leave with a senior-level design lens—patterns for reliability, performance, security, and LLM evaluation—to refactor agents and prompts before they hit real traffic.
- Production-oriented agentic architecture recommendations
- Evaluation and observability considerations for LLM agents
- Security and performance design checklist aligned to the skill sections
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Primary shelf is Build/agent-tooling because the skill centers on designing agentic systems and prompt-engineering craft while you assemble agent products. agent-tooling fits procedural prompts, tool orchestration, and senior-level patterns rather than a single UI or API endpoint task.
Where it fits
Refactor a multi-tool agent graph with clearer prompt boundaries and batching before adding new MCP tools.
Set latency and caching strategy for a real-time agent endpoint before public beta.
Define what to log and measure so prompt regressions show up in dashboards, not support tickets.
How it compares
A design-and-eval playbook for agentic systems—not a lightweight single-task generator like a one-shot PR description skill.
Common Questions / FAQ
Who is senior-prompt-engineer for?
Indie and small-team builders treating LLM agents as products who need senior prompt-engineering and agentic architecture judgment, not just chat templates.
When should I use senior-prompt-engineer?
While building agent-tooling (prompt/tool graphs), before Ship hardening (security, perf targets), and when Operate needs observability and eval loops for LLM behavior.
Is senior-prompt-engineer safe to install?
It is advisory documentation; review the Security Audits panel on this Prism page and avoid letting agents apply production changes without your review.
SKILL.md
READMESKILL.md - Senior Prompt Engineer
# Agentic System Design ## Overview World-class agentic system design for senior prompt 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 # Llm Evaluation Frameworks ## Overview World-class llm evaluation frameworks for senior prompt 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 # Prompt Engineering Patterns ## Overview World-class prompt engineering patterns for senior prompt 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 retrie