aws-agentic-ai

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation commands and documentation from official AWS sources and the aws-samples GitHub organization, which is consistent with its role as a developer resource.
  • [COMMAND_EXECUTION]: The skill provides tools for executing AWS CLI commands (aws bedrock-agentcore, aws secretsmanager, aws s3, aws sts). These capabilities are necessary for the skill's primary purpose of deploying and managing agentic infrastructure on AWS.
  • [INDIRECT_PROMPT_INJECTION]: The skill features ingestion points where it reads external documentation (mcp__acdocs__fetch_agentcore_doc) and registry records (aws bedrock-agentcore search-registry-records). While these provide an attack surface for untrusted data, the skill's instructions focus on providing expert guidance to the user, and the risk is inherent to its function as a technical assistant.
  • Ingestion points: Documentation retrieval tools and registry search outputs.
  • Boundary markers: None explicitly defined in the static instruction set.
  • Capability inventory: Significant cloud resource management capabilities via the AWS CLI Bash tools.
  • Sanitization: Standard CLI output parsing is expected, though not explicitly detailed in the markdown instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 12:39 AM
Security Audit — agent-trust-hub — aws-agentic-ai