superplane-app-builder

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill facilitates the creation of workflows that ingest data from external, potentially untrusted sources (such as GitHub PR comments, issue bodies, or Sentry error metadata) and use that data in powerful actions like shell command execution or HTTP requests. This configuration creates an indirect prompt injection surface.
  • Ingestion points: Data enters the agent's context through GitHub triggers (references/github.md) and Sentry triggers (references/sentry.md).
  • Boundary markers: The skill does not provide specific instructions or markers to delimit external content or warn the underlying execution engine to ignore embedded instructions within processed payloads.
  • Capability inventory: The generated workflows have access to shell execution via daytona.executeCommand (references/daytona.md), ssh commands (references/components-and-triggers.md), and arbitrary http requests (references/components-and-triggers.md).
  • Sanitization: While the skill recommends "hardened" command templates and shell quoting practices in references/daytona.md, it lacks explicit requirements for sanitizing or validating external strings before they are interpolated into executable expressions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 05:27 PM
Security Audit — agent-trust-hub — superplane-app-builder