Back to skill

Security audit

odq-crypto-data

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed pay-per-request crypto data skill, with the main caution that users should approve any paid call or wallet-address lookup before it runs.

Install only if you are comfortable with an x402-compatible client spending small amounts of USDC on Base for each request. Require confirmation before paid calls, especially loops or watchlists, and avoid querying wallet addresses unless the address owner has consented or the privacy tradeoff is acceptable.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

External Script Fetching

High
Category
Supply Chain
Confidence
90% confidence
Finding

Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.

Content

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

Quick start (curl)

Any x402-compatible client works (Bankr terminal x402.fetch, x402-axios, x402 SDKs). Manual curl flow:

bash
URL="$BASE/market-signal?coin=bitcoin"

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs use of pay-per-request endpoints with wallet-based USDC settlement but does not clearly warn that invoking the APIs can automatically spend funds. In an agent setting, this creates a real risk of unintended micropayments or repeated charges if the tool is called automatically or in loops, especially because the user may not realize execution has a monetary consequence.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The manifest says the skill is for querying 6 pay-per-request crypto-data APIs and lists market signals, DEX scans, token safety, funding heatmap, perp regime, and wallet snapshots. This reference file documents a seventh endpoint, wallet-watch, which means the documented behavior is broader than the manifest’s explicit six-endpoint scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The wallet-watch endpoint queries third-party infrastructure for wallet balances and recent transactions, which can disclose user-linked addresses and access patterns to external services without an explicit privacy warning. In a crypto context, wallet addresses can be sensitive metadata because they reveal holdings, transaction history, and behavioral associations even if they are publicly visible on-chain.

Content

No source excerpt is available for this finding.