customerio-reference-architecture

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill architecture includes a webhook handler that represents an indirect prompt injection surface.
  • Ingestion points: SKILL.md defines a CustomerIOWebhooks class that processes req.body.events.
  • Boundary markers: No explicit boundaries or instructions to ignore embedded content are provided in the templates.
  • Capability inventory: The skill uses Bash(gh:) and Bash(curl:) tools.
  • Sanitization: The provided code emits event data directly without validation or sanitization.
  • [COMMAND_EXECUTION]: The skill defines high-privilege tool access.
  • Permission for Bash(gh:) and Bash(curl:) is requested, which grants significant control over the environment and external repositories. This is consistent with the skill's purpose but increases the impact of other vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 01:03 AM