To extract action items from a meeting transcript, which prompt approach is MOST effective?
Select an answer to reveal the explanation.
Short Explanation
When you need to pull action items from a transcript, give Claude a precise spec: fields, format, and missing-data rule. Vague prompts get vague outputs.
Full Explanation
For information extraction tasks like mining action items from transcripts, precise prompt specification is essential. Specify: what to extract (action items vs. decisions vs. questions), the output format (numbered list, JSON, table), required fields (owner, task, due date), how to handle edge cases (no due date → 'TBD'), and whether to include items from all speakers or specific ones. This specification eliminates Claude's need to guess at your requirements. The approach that specifies the exact output format, including owner name, task description, due date if mentioned, and 'TBD' when no due date is given, is the most effective. The prompt asking Claude to 'read this transcript and tell me things' is maximally vague. The prompt using temperature=1.0 to generate the most creative interpretation introduces randomness — not appropriate for precise extraction (use temperature=0). The prompt sending the transcript without any instructions and hoping Claude extracts action items provides no guidance.