aws-well-architected-framework-review

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). SKILL.md Step 1/4 instructs the agent to use the user-provided “Workload description” and whatever “code packages/directories” (plus any verbatim architecture details when no code exists) as evidence for the WA evaluation, meaning outsider-authored free text can directly enter the model context to drive subsequent analysis.