Back to skill

Security audit

NetIntel

Security checks for vulnerabilities and agentic risk

Overview

The skill is transparent about paid NetIntel lookups, but it documents unpinned wallet/payment setup commands that may handle funds or private keys.

Install only if you are comfortable with pay-per-call external lookups. Use a dedicated low-balance wallet, set per-call and session limits, require explicit approval for paid calls, and avoid running unpinned `npx ...@latest` wallet or MCP commands with private keys unless you have independently reviewed and pinned the package version.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 11)May include surrounding context.

md
requires:
      anyBins:
        - npx
        - curl
    envVars:
      - name: EVM_PRIVATE_KEY
        required: false

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
90% confidence
Finding

The skill states that payments are 'usually automatic' and that some wallet integrations pay every 402 they meet without asking. Even though the document later advises obtaining user authorization, the default behavior described here can enable autonomous spending if an agent follows the flow naively or if prompts trigger repeated paid lookups.

Content

Scanner excerpt · SKILL.md (reported line 39)May include surrounding context.

md
## Payment, authorization and spending limits — read before any paid call

- **Payment is usually automatic.** The `netintel` MCP server with a key, AgentCash and the
  Coinbase Agentic Wallet pay every 402 they meet without asking; Coinbase Wallet MCP asks the
  user to approve each payment. The prices in this file are what a call costs; they are
  **not** a spending cap.
- **Get the user's authorization before paid calls.** Tell the user what you intend to call

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 58)May include surrounding context.

md
## How to call and pay

Every endpoint is a plain HTTPS URL, so this skill works on its own with whichever wallet the
agent already has. Use the first option that applies. **Never ask the user to paste a private
key into the conversation.** Send the header `X-NetIntel-Via: skill` with these calls (as below):
it only tells NetIntel which integration a call came from and changes nothing about the call or
its price.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

The skill instructs users to execute npx awal@latest, which fetches and runs code from the registry at invocation time without pinning an immutable version. If the package or a dependency is compromised, or a breaking release is published, the agent could execute attacker-controlled code with wallet and environment access.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

This line repeats use of npx awal@latest for a paid POST flow, again causing unreviewed remote code to be downloaded and executed at runtime. Because the command is tied to payment operations, compromise could lead to wallet misuse, payment redirection, or secret theft.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
91% confidence
Finding

The setup command npx -y agentcash@latest installs and runs a wallet/payment tool without version pinning. That creates a supply-chain execution risk, especially dangerous because the tool may handle funds and authenticated payment requests.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
93% confidence
Finding

This login/setup path also uses npx awal@latest, exposing users to the same supply-chain risk during wallet authentication. A compromised package could capture credentials, session tokens, or manipulate subsequent payment behavior.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The command npx -y netintel-mcp is particularly sensitive because it is shown alongside passing EVM_PRIVATE_KEY=0x… into the environment. Running an unpinned package in a context where a private key is available could result in direct key exfiltration and unauthorized asset transfers if the package or its dependencies are compromised.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.