skills/openclaw/skills/shadow-number/Gen Agent Trust Hub

shadow-number

Fail

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill requires a sensitive environment variable SHADOW_WALLET_KEY, which is used as a private key for a crypto wallet to handle payments.
  • Evidence: The YAML frontmatter in SKILL.md identifies SHADOW_WALLET_KEY as a required environment variable.
  • [EXTERNAL_DOWNLOADS]: The skill relies on an external API hosted on an unverified third-party domain, which is not recognized as a trusted organization or well-known service.
  • Evidence: SKILL.md directs all network requests to https://extraordinary-charisma-production.up.railway.app for purchasing numbers and polling for OTP codes.
  • [DATA_EXFILTRATION]: Sensitive wallet credentials are used to interact with a remote, untrusted server for the purpose of executing automated financial transactions.
  • Evidence: SKILL.md documentation states the skill automatically handles USDC payments on the Base network using the provided key when the API returns an HTTP 402 status code.
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes and acts upon data received from an external, untrusted source.
  • Ingestion points: Data is retrieved from the /api/smspva/otp/{orderId} endpoint.
  • Boundary markers: None are present to distinguish between data and potential instructions.
  • Capability inventory: The agent uses the data to navigate browser windows and submit web forms.
  • Sanitization: No sanitization or validation of the received code is mentioned in the skill instructions.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 26, 2026, 04:17 PM