senior-prompt-engineer

This skill should be used when the user asks to "optimize prompts", "design prompt templates", "evaluate LLM outputs", "build agentic systems", "implement RAG", "create few-shot examples", "analyze token usage", or "design AI workflows". Use for prompt engineering patterns, LLM evaluation frameworks, agent architectures, and structured output design.

1.18x

Quality

53%

Does it follow best practices?

Impact

91%

1.18x

Average score across 6 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./engineering-team/senior-prompt-engineer/SKILL.md

Evaluation results

95%

Customer Support Prompt Cost Reduction

Prompt optimization workflow

Criteria

Without context

With context

Baseline analysis run

91%

100%

Baseline saved to file

100%

GPT-4 model specified

100%

Optimized version produced

100%

Optimizer script used for optimization

25%

37%

Comparison report generated

100%

Compare flag used

60%

100%

Issue identification step

100%

Optimization pattern applied

100%

Workflow log documents steps

100%

Knowledge Base Q&A System Quality Audit

RAG system quality evaluation

Criteria

Without context

With context

rag_evaluator.py used

100%

JSON output file produced

100%

Verbose flag used

100%

Output flag used

100%

Retrieval metrics present

100%

Generation metrics present

100%

Recommendations documented

100%

Evaluation commands logged

100%

Issues section in report

100%

10%

Research Agent Architecture Review

Agent workflow design and documentation

Criteria

Without context

With context

--validate flag used

100%

ASCII visualize flag used

100%

Mermaid format flag used

100%

--format mermaid flag used

100%

--estimate-cost flag used

100%

--runs 50 specified

100%

All three analyses referenced in summary

100%

Script output used (not reimplemented)

100%

--output flag used

100%

88%

22%

Shipping Complaint Extraction Prompt

Few-shot example design workflow

Criteria

Without context

With context

Example count 3-5

100%

Simple/typical case

100%

Diverse example types

80%

100%

Consistent Input/Output format

66%

100%

Task description first

100%

Output format specified

100%

--extract-examples used

100%

--validate-examples attempted

examples.json produced

100%

85%

20%

Contract Metadata Extraction Prompt

Structured output prompt design workflow

Criteria

Without context

With context

Schema with typed fields

100%

Schema is valid JSON

100%

Schema in prompt

100%

ONLY JSON enforcement

100%

Start/end delimiter constraint

100%

No markdown/explanation

100%

--validate-schema attempted

--analyze run on prompt

100%

Required fields marked

100%

80%

26%

Prompt Audit and Strategy Recommendation

Token cost estimation and pattern selection

Criteria

Without context

With context

--tokens --model gpt-4 used

--json flag used

100%

Token count stored in file

70%

100%

--analyze run on prompt

100%

Two or more named patterns

100%

Pattern recommendation given

100%

Cost or token reasoning

100%

Issues identified from analysis

70%

100%

Repository: alirezarezvani/claude-skills
Commit: 967fe01

Evaluated: 25 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Customer Support Prompt Cost Reduction Knowledge Base Q&A System Quality Audit Research Agent Architecture Review Shipping Complaint Extraction Prompt Contract Metadata Extraction Prompt Prompt Audit and Strategy Recommendation

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.