skills/refly-ai/refly-skills/stripe/Gen Agent Trust Hub

stripe

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • COMMAND_EXECUTION (HIGH): The execution steps in SKILL.md demonstrate the construction of shell commands using user-provided input fields such as "invoice_description". Direct interpolation of these fields into a single-quoted shell string for the "refly" CLI enables arbitrary command execution if an attacker provides inputs containing shell metacharacters.
  • Ingestion points: Untrusted user input via the "customer_name", "customer_email", and "invoice_description" fields defined in SKILL.md.
  • Boundary markers: Absent; inputs are directly embedded into the shell command string.
  • Capability inventory: Execution of shell commands via the "refly" CLI, which triggers external payment and billing workflows.
  • Sanitization: Absent; no escaping of shell metacharacters is required or implemented in the skill instructions.
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill requires the installation of an opaque package ID ("skp-mfyz61qqatnfv05xj9nruiom") and executes a workflow on a third-party platform ("refly.ai") via a specific workflow ID. The logic within these remote components is not transparent, and the author ("chizblank") is not a trusted source.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 06:28 AM