builder
Prompt engineering
//
Prompt review and hardeningleaf
Adversarial review of an LLM prompt with named failure modes and a hardened rewrite.
Chain-of-thought reviewleaf
Review a CoT prompt for reasoning quality, not just final-answer accuracy.
Agent tool design (function-calling)leaf
Design tools (function-calling) for an LLM agent so the agent succeeds without manual handholding.