Solana Sniper Bot

• Outcome: Review
• Latest audit: May 28, 2026 at 12:29 PM UTC
• Version:
  1.0.0

Overview

This skill is a disclosed Solana trading bot, but it can use a wallet private key to attempt autonomous live trades without a confirmation or dry-run safety gate.

Review carefully before installing. Use only a dedicated low-balance wallet, assume trades can lose all allocated funds, do not rely on the documented stop-loss or auto-sell protections without implementing and testing them, and prefer a dry-run or manual-approval workflow before allowing live transactions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 92% confidence
• Finding: The skill requires sensitive environment secrets, writes files during deployment, and performs network actions, but does not declare explicit permissions. In an agent ecosystem, this creates a capability mismatch that can lead users or orchestrators to grant or execute powerful behavior without clear visibility, especially dangerous here because the skill handles a Solana private key and can place real trades.

Intent-Code Divergence

High

• Confidence: 98% confidence
• Finding: The function claims to execute a Jupiter swap, but it never attaches the generated signature to the deserialized transaction before sending it. As written, the bot may submit an unsigned or improperly signed transaction while using skipPreflight, causing failed trades, inconsistent accounting, and potentially unsafe assumptions that orders were executed when they were not.

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The invocation text is broad enough to match generic requests about Solana trading, token monitoring, or building bots, which increases the chance the skill is auto-selected in contexts where the user did not intend autonomous live trading. In this skill, overbroad triggering is more dangerous than usual because activation can expose wallet secrets and lead to automated financial transactions on highly risky assets.

Missing User Warnings

High

• Confidence: 95% confidence
• Finding: The bot automatically buys newly detected tokens based on heuristic and LLM scoring with no user confirmation, rate limiting beyond position count, or final approval step. In this skill context, that is especially dangerous because it controls a real wallet and targets highly adversarial memecoin launches where bad classifications can quickly lead to financial loss.

VirusTotal

65/65 vendors flagged this skill as clean.