support-cases
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection Analysis: This skill contains an attack surface for indirect prompt injection through its ingestion of external data.
- Ingestion points: The skill retrieves communication threads from AWS Support cases via the
aws support describe-communicationscommand as outlined inSKILL.md. - Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between its instructions and the potentially untrusted text within the case communications.
- Capability inventory: The skill primarily uses read-only commands (
describe-cases,describe-communications) inSKILL.md, though the underlying agent may possess broader capabilities that could be targeted by an injection. - Sanitization: There are no specific steps mentioned for sanitizing or filtering the content of the support case messages before analysis.
- Command Execution via AWS CLI: The skill provides instructions for using the AWS CLI to interact with the AWS Support API.
- The
support:DescribeCasesandsupport:DescribeCommunicationsoperations are used to search and retrieve incident history. This is consistent with the skill's primary purpose and uses standard vendor tooling.
Audit Metadata