Does whispering to agents in docs help?
Summary
A technical blog post exploring whether whispering to agents in documentation helps. The author runs experiments with agent-focused prompts and finds that explicit recommendations strongly influence agent decisions, often regardless of other contextual changes. The results suggest that well-written, human-friendly docs with clear examples are more effective than labeling content for agents, and the author provides practical guidelines for making docs usable by both humans and AI agents.