ecs-observability

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No malicious instruction overrides or behavior bypass patterns were detected. Instructional language uses terms like 'IMPORTANT' and 'non-negotiable' solely to emphasize technical accuracy regarding AWS launch types.
  • [DATA_EXFILTRATION]: No evidence of data exposure or exfiltration. The skill does not access sensitive local files or perform unauthorized network operations. CLI commands mentioned (e.g., 'aws ecs list-account-settings') are standard tools for its stated purpose.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns or unverifiable dependencies found. The skill refers to official AWS container images (e.g., 'public.ecr.aws/aws-observability/aws-otel-collector') and documented configuration parameters.
  • [COMMAND_EXECUTION]: Shell commands listed in the documentation are for administrative auditing and configuration of ECS clusters, aligned with the skill's primary advisory role.
  • [EXTERNAL_DOWNLOADS]: All external references point to trusted domains, including official AWS documentation ('docs.aws.amazon.com'), AWS blogs ('aws.amazon.com/blogs'), and the official AWS OpenTelemetry project site ('aws-otel.github.io').
  • [SAFE]: No obfuscation, persistence mechanisms, or privilege escalation patterns were identified across the 6 analyzed files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 11:00 AM
Security Audit — agent-trust-hub — ecs-observability