aws-bedrock-agentcore-skill
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as an educational and operational guide for Amazon Bedrock and AgentCore services. Its content is oriented toward promoting security, such as explicitly warning against overly broad IAM permissions and documenting the use of Bedrock Guardrails to mitigate prompt injection.
- [SAFE]: All code snippets are provided as copy-paste starters for legitimate development tasks. They demonstrate secure coding patterns, such as using context managers for session cleanup, explicit region configuration to avoid misconfiguration, and the use of environment variables for secrets.
- [SAFE]: External references are limited to official AWS documentation (docs.aws.amazon.com), AWS-maintained GitHub organizations (strands-agents, aws-ia, awslabs), and standard package registries (PyPI, NPM). No suspicious or unknown third-party domains were identified.
- [SAFE]: No patterns of obfuscation, credential exfiltration, or unauthorized command execution were found. The skill emphasizes the 'Assume-Malicious' posture by documenting threat models like the confused deputy attack and how to prevent it using IAM conditions.
Audit Metadata