langgraph-agents

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is an educational resource from a trusted vendor providing standard implementation patterns for AI agents. It recommends best practices such as using Amazon Bedrock Guardrails for safety.
  • [EXTERNAL_DOWNLOADS]: The documentation references official AWS resources on GitHub and well-known Python libraries (langgraph, langchain-aws, boto3) for building agentic workflows.
  • [PROMPT_INJECTION]: The skill demonstrates processing external game state data for agent decision-making.
  • Ingestion points: The lambda_handler and agent.invoke code samples in SKILL.md.
  • Boundary markers: Prompts use standard role-based message separation (system/user).
  • Capability inventory: The provided tools are restricted to math and game-specific tactical logic.
  • Sanitization: The tutorial describes patterns for summarizing input and parsing output before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 06:30 PM
Security Audit — agent-trust-hub — langgraph-agents