ClawHub
Back to skill

Security audit

SushiSwap API

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only SushiSwap API skill whose financial transaction-data capability is disclosed, with no hidden execution or persistence found.

Safe to install as an API reference skill, but treat generated swap transactions as financial actions. Before signing anything in a wallet, verify the chain, token addresses, amount, recipient, slippage, fee settings, referrer, target contract, and transaction value; prefer the production Sushi API server rather than the included SwaggerHub mock server.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Vague Triggers

Medium
Confidence
82% confidence
Finding
This file is a YAML manifest/spec, so SQP-1 applies. The descriptions 'generates a swap' and 'generate a swap transaction' are broad and do not clarify operational boundaries, exclusions, or negative examples, which can make downstream agents invoke the capability too readily when users discuss swaps in general.

Missing User Warnings

High
Confidence
93% confidence
Finding
For markdown/manifest-style skill descriptions, warnings should be present when behavior can affect user data or system integrity. The /swap endpoint requires sender and optional recipient addresses and returns a transaction object with calldata and value, but the spec does not warn that using it can facilitate on-chain asset transfers or disclose wallet-related data to the API.

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Content
version: 7.0.0

  # API description. Arbitrary text in CommonMark or HTML.
  description: Documentation for interacting with the Sushi API

servers:
  - description: Sushi API
Confidence
80% confidence
Finding
description:; description:; description:; description:; description:; description:; description:; description:; parameters:; description:; description:; description:; description:; description:; descr

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.