Drop Noor into Claude and get a senior prompt engineer who designs, tests and versions production prompts with an eval harness, not vibes.
Noor engineers production prompts: prompt patterns (zero and few-shot, chain-of-thought, decomposition), system-prompt design, structured/JSON output with schema enforcement, retrieval-aware and tool-calling prompts, prompt evaluation and testing (eval sets, regression tests, LLM-as-judge, rubrics), prompt versioning, token/cost and latency optimization, and guardrails and jailbreak resistance. Always test prompts against an eval set in dev/staging, version prompts, add guardrails, and review quality and safety before shipping.
What you get
-
→Prompt patterns: few-shot, chain-of-thought, decomposition
-
→Structured/JSON output and schema enforcement
-
→Prompt evaluation: eval sets, LLM-as-judge, rubrics, regression
-
→Versioning, cost/latency optimization and guardrails
📄 noor-prompt-engineer.skill
Under 2 min install
Works with Claude, ChatGPT & any AI chat
How to install
Download the .skill package → open Claude → paste SKILL.md into your Project Instructions or system prompt → describe your requirement → Noor builds the answer. Includes a full worked example so you see exactly what you get.