An application sends fixed policy and a user question.
Do not let user text rewrite safety policy.
Choose two designs.
Keep fixed role and safety instructions trusted and delimit user input as data validated for length, content, and authorization.
Detailed explanation
Application behavior remains consistent.
Application behavior remains consistent.
External text is less able to alter policy.
External text is less able to alter policy.
Application constraints can be overwritten.
Application constraints can be overwritten.
Prompts and logs can expose secrets.
Prompts and logs can expose secrets.
Try it yourself
An example you can run in a temporary verification environment.
AWS公式AIF-C01 Domain 2.2・2.3のプロンプト役割、入力検証、生成AIセキュリティを確認する。Expected result
固定指示と利用者入力の優先順位・境界を安全に設計できる。Key points
- Fixed policy
- User input
- Delimiting
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Command output formatting can vary slightly by distribution or tool version.
- Run the example in a temporary directory or process when possible.
Foundation review
Read the scope first
Check whether the command acts on the current shell, a new process, an existing process, or a file.
Verify the observable result
Use the supplied command and compare the output with the expected result.