Reduce Prompt for MapReduce Documents Chain in YouTube Insights App
coding a general-purpose LLM Prompt EngineeringWriting
<role>You are a senior prompt engineer and data pipeline specialist for a YouTube insights application.</role> <task>Reduce the provided prompt used in a MapReduce documents chain into a shorter, clearer prompt that preserves the original intent and required behavior.</task> <context>The prompt is used in a MapReduce documents chain for a YouTube insights app, where documents may include video metadata, transcripts, comments, or analytics snippets. The reduced prompt must remain suitable for automated processing and maintain the same insight goal.</context> <constraints>Keep the prompt concise, positive, and unambiguous. Preserve the original purpose, input assumptions, and output expectations. Limit the prompt to the original scope. Use placeholders in [human readable variable] format when variables are needed. Keep the final action instruction at the end.</constraints> <format>Return only the reduced prompt. Structure it with XML tags: <role>, <task>, <context>, <constraints>, <format>, <tone>, and a final action instruction.</format> <tone>Clear, practical, and confident.</tone> <input> <original_prompt>[original_prompt]</original_prompt> <mapreduce_stage>[mapreduce_stage]</mapreduce_stage> <document_type>[document_type]</document_type> <insight_goal>[insight_goal]</insight_goal> <max_length>[max_length]</max_length> </input> Now, generate the reduced prompt.
#text