ClawHub
Back to skill

Security audit

VeraData — LATAM Compliance API

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed remote compliance API skill with sensitive-data and micropayment considerations, but the behavior is coherent with its stated purpose and no hidden or destructive behavior was found.

Install only if you are comfortable sending sanctions, KYB, registry, or rates queries to VeraData's hosted service and allowing x402 USDC micropayments under your agent's controls. Use trial mode or set explicit approval and budget limits before enabling paid calls or the remote MCP server, and avoid submitting regulated personal or business data unless your organization approves the provider's retention and audit model.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (12)

Vague Triggers

Medium
Confidence
73% confidence
Finding
The skill uses very broad invocation language ('use this skill whenever a task involves...'), which can cause agents to route sensitive compliance, KYC/KYB, sanctions, and registry tasks to this external service by default. In this context, overbroad triggering increases the chance of unnecessary third-party data disclosure and autonomous paid actions for tasks that may not require this provider.

External Transmission

Medium
Category
Data Exfiltration
Content
author: teodorofodocrispin-cmyk
  version: "2.3.2"
  base_url: https://api.veradata.dev
  health: https://api.veradata.dev/health
  x402_discovery: https://api.veradata.dev/.well-known/x402
  llms_txt: https://api.veradata.dev/llms.txt
  openapi: https://api.veradata.dev/openapi.json
Confidence
96% confidence
Finding
The skill is explicitly designed to transmit user-supplied compliance and identity-related data to an external API endpoint. In this context, the transmission is functional rather than covert, but it still creates a real privacy and security risk because sanctions/KYB/registry queries may contain personal or corporate sensitive information sent off-platform.

External Transmission

Medium
Category
Data Exfiltration
Content
version: "2.3.2"
  base_url: https://api.veradata.dev
  health: https://api.veradata.dev/health
  x402_discovery: https://api.veradata.dev/.well-known/x402
  llms_txt: https://api.veradata.dev/llms.txt
  openapi: https://api.veradata.dev/openapi.json
  mcp: https://api.veradata.dev/mcp
Confidence
96% confidence
Finding
This line documents another external service endpoint under the same API domain, reinforcing that the skill depends on outbound network access. Because the service handles compliance and sanctions screening, sending requests externally can expose sensitive names, entities, and related metadata even if the provider claims limited retention.

External Transmission

Medium
Category
Data Exfiltration
Content
base_url: https://api.veradata.dev
  health: https://api.veradata.dev/health
  x402_discovery: https://api.veradata.dev/.well-known/x402
  llms_txt: https://api.veradata.dev/llms.txt
  openapi: https://api.veradata.dev/openapi.json
  mcp: https://api.veradata.dev/mcp
  payment_networks:
Confidence
96% confidence
Finding
The documented openapi/discovery resources indicate multiple external touchpoints that an agent may query automatically. While normal for an API-backed skill, this expands the external attack and privacy surface by encouraging autonomous retrieval of remote manifests and schemas.

External Transmission

Medium
Category
Data Exfiltration
Content
health: https://api.veradata.dev/health
  x402_discovery: https://api.veradata.dev/.well-known/x402
  llms_txt: https://api.veradata.dev/llms.txt
  openapi: https://api.veradata.dev/openapi.json
  mcp: https://api.veradata.dev/mcp
  payment_networks:
    base: "eip155:8453 — USDC 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"
Confidence
95% confidence
Finding
The skill instructs agents to interact with external infrastructure for discovery and operation, which is a real external-transmission behavior. In a compliance-focused skill, even support or discovery calls can leak usage patterns, IP information, and operational metadata to the vendor.

External Transmission

Medium
Category
Data Exfiltration
Content
x402_discovery: https://api.veradata.dev/.well-known/x402
  llms_txt: https://api.veradata.dev/llms.txt
  openapi: https://api.veradata.dev/openapi.json
  mcp: https://api.veradata.dev/mcp
  payment_networks:
    base: "eip155:8453 — USDC 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"
    solana: "solana:5eykt4UsFv8P8NJdTREpY1vzqKqZKvdp — USDC EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v"
Confidence
95% confidence
Finding
This content continues the external-service and payment-network configuration, showing that operation depends on outside systems. The main risk is not the existence of URLs themselves, but that agents may autonomously connect to third-party services and payment flows in compliance-sensitive contexts.

External Transmission

Medium
Category
Data Exfiltration
Content
### 5. Operator Verification

- **GitHub:** https://github.com/teodorofodocrispin-cmyk
- **Health:** https://api.veradata.dev/health
- **x402 discovery:** https://api.veradata.dev/.well-known/x402
- **Contact:** teodorofodocrispin@gmail.com
- **ERC-8299 conformance:** https://github.com/babyblueviper1/preaction-governance-conformance
Confidence
92% confidence
Finding
The operator verification section references external resources that agents or users may consult, confirming dependence on third-party infrastructure. This is not inherently malicious, but in aggregate it contributes to data egress and trust-on-first-use risk if agents automatically fetch and trust remote verification material.

External Transmission

Medium
Category
Data Exfiltration
Content
- **GitHub:** https://github.com/teodorofodocrispin-cmyk
- **Health:** https://api.veradata.dev/health
- **x402 discovery:** https://api.veradata.dev/.well-known/x402
- **Contact:** teodorofodocrispin@gmail.com
- **ERC-8299 conformance:** https://github.com/babyblueviper1/preaction-governance-conformance
Confidence
92% confidence
Finding
Additional external references in the operator verification section expand the set of remote locations an agent may contact. In a security review, this is a true external-transmission concern because the skill encourages remote dependency checks and service interaction outside the local trust boundary.

External Transmission

Medium
Category
Data Exfiltration
Content
{
  "mcpServers": {
    "veradata": {
      "url": "https://api.veradata.dev/mcp",
      "transport": "http"
    }
  }
Confidence
97% confidence
Finding
The MCP integration exposes a remote tool server that an agent may connect to over HTTP transport, potentially enabling broad remote access to compliance-related functions. In this context, MCP access is more dangerous than a simple REST call because it can widen the agent's operational surface and make autonomous external tool use easier and more persistent.

External Transmission

Medium
Category
Data Exfiltration
Content
## Resources

- Health check: https://api.veradata.dev/health
- x402 discovery: https://api.veradata.dev/.well-known/x402
- OpenAPI: https://api.veradata.dev/openapi.json
- llms.txt: https://api.veradata.dev/llms.txt
Confidence
93% confidence
Finding
The resources section encourages contact with external health/discovery endpoints, which is a genuine outbound network behavior. Though common for API skills, it still matters in regulated workflows because it extends the external footprint and may reveal operational metadata.

External Transmission

Medium
Category
Data Exfiltration
Content
## Resources

- Health check: https://api.veradata.dev/health
- x402 discovery: https://api.veradata.dev/.well-known/x402
- OpenAPI: https://api.veradata.dev/openapi.json
- llms.txt: https://api.veradata.dev/llms.txt
- A2A manifest: https://api.veradata.dev/.well-known/a2a-agent.json
Confidence
93% confidence
Finding
This line adds further external references that may be queried by agents for manifests and metadata. The danger is contextual: this skill handles compliance and identity-adjacent workflows, so unnecessary remote lookups can increase exposure without advancing the user task.

External Transmission

Medium
Category
Data Exfiltration
Content
- Health check: https://api.veradata.dev/health
- x402 discovery: https://api.veradata.dev/.well-known/x402
- OpenAPI: https://api.veradata.dev/openapi.json
- llms.txt: https://api.veradata.dev/llms.txt
- A2A manifest: https://api.veradata.dev/.well-known/a2a-agent.json
- GitHub (public): https://github.com/teodorofodocrispin-cmyk/veradata-public
Confidence
93% confidence
Finding
The public resource links confirm multiple outbound destinations associated with this skill. While not evidence of malice, they form part of a true data egress and supply-chain trust surface that is more sensitive here because the skill is intended for sanctions, KYB, and regulatory screening.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.