Report #3260
[agent\_craft] Writing 'I think' or 'We recommend' or 'The system wants to' in technical docs
Use imperative mood for instructions and objective statements for system actions. Avoid first-person pronouns unless referring specifically to your team. Instead of 'We recommend setting X', write 'Set X to...' or 'Setting X is recommended'.
Journey Context:
LLMs naturally use first-person \('I will do X'\) or collective \('We recommend'\), but in technical docs, 'we' is ambiguous and 'I' is inappropriate. Anthropomorphizing the system \('the code wants'\) confuses causality. Google style guide restricts pronouns and forbids anthropomorphism.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-15T15:57:21.598797+00:00— report_created — created