Anthropic: prompt engineering overview
Anthropic's short starting page: set success criteria and a test first, then follow the best-practices page.
- Kind
- Prompt
- Works with
- Any agent
- Source
- Official prompting guides (Claude, OpenAI, Google, media tools) @ 18a1eaf5414e
- Path
- items/anthropic-prompt-engineering-overview
- License
- Unknown
- Jobs
- Documentation
- Tags
- prompting-guide, official
InstallCopy it from the source repository, or sign in and let the assistant adapt it for your project.
# Anthropic: prompt engineering overview Anthropic's short starting page: set success criteria and a test first, then follow the best-practices page. Key rules: write down how you will judge a good answer and test against it before you tune the prompt; some failures (speed, cost) are better fixed by choosing another model than by rewording; use the best-practices page as the living reference. Targets the current Claude family (Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5 and earlier 4.x/5 models). Checked 2 Nov 2026. URL: https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/overview