aws-lambda

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes several local binaries including aws, jq, and zip. These are used as intended for cloud resource management and data processing.
  • [PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface because it processes untrusted data from remote sources.
  • Ingestion points: Data enters the agent context through the output of aws lambda invoke (stored in /tmp/lambda-out.json) and the results of aws logs tail.
  • Boundary markers: No delimiters or explicit instructions to ignore embedded content are provided in the command sequence.
  • Capability inventory: The skill possesses high-privilege capabilities including the ability to invoke functions, update function code, and modify environment configurations.
  • Sanitization: There is no evidence of sanitization or validation of the external content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 05:13 AM