aws-cloudwatch

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides examples using the aws CLI tool to query logs, metrics, and CloudTrail events. These are standard administrative operations.
  • [PROMPT_INJECTION]: The skill demonstrates constructing CloudWatch Insights and CloudTrail Lake queries using user-provided inputs such as log group names and query strings (e.g., in SKILL.md).
  • Ingestion points: logGroupNames, queryString, InstanceId, AlarmName, QueryStatement in SKILL.md.
  • Boundary markers: Absent.
  • Capability inventory: boto3 client operations and aws CLI subshell executions.
  • Sanitization: None shown in snippets. While these represent a surface for indirect prompt injection, the behavior is consistent with the primary purpose of a monitoring utility.
  • [SAFE]: No malicious obfuscation, credential theft, or unauthorized network operations were detected. All external references involve standard AWS service interactions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 07:12 AM
Security Audit — agent-trust-hub — aws-cloudwatch