serverless-development

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill demonstrates handlers for API Gateway and SQS that ingest untrusted user input via event.body and record.body. While the examples use standard parsing, they do not include explicit schema validation or sanitization before performing database operations, representing a potential surface for injection if downstream consumers do not perform validation.
  • Evidence: SKILL.md (apiHandler, sqsHandler).
  • Capability Inventory: Performs database writes (db.user.create) and reads (db.user.findMany) based on inputs.
  • Sanitization: Uses standard JSON.parse but lacks content-specific filtering for properties stored in the database.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 08:36 PM
Security Audit — agent-trust-hub — serverless-development