Base Trader

• Outcome: Review
• Latest audit: Jul 10, 2026 · 2:04 AM UTC
• Version: 1.1.1

Overview

This skill is openly for crypto trading, but it can affect real wallet funds through broad prompts and autonomous trading workflows without clear per-trade consent controls.

Install only if you are comfortable giving an agent workflows that can trade real crypto. Use a dedicated low-balance wallet, inspect the separate Bankr skill and its permissions, require manual confirmation for every buy/sell/order, and avoid enabling cron, heartbeat, DCA, leverage, or standing orders until spending limits and cancellation steps are clear.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (11)

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The README states the skill triggers automatically on broad trading-related queries such as "trade," "buy," "sell," and portfolio questions. In a skill capable of autonomous crypto trading, overly broad activation increases the chance of unintended invocation and accidental execution of wallet-affecting actions, especially when normal conversational queries may be interpreted as trade instructions.

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The usage section shows natural-language examples like "Buy $20 of TOKEN on Base" and says the skill triggers automatically, but it does not prominently warn that these interactions may execute real trades using the user's Bankr wallet and funds. In the context of an autonomous trading skill connected to a live wallet, insufficient disclosure materially increases the risk of users triggering irreversible financial transactions without fully informed consent.

Vague Triggers

High

• Confidence: 96% confidence
• Finding: The trigger description is extremely broad, matching generic terms like "buy," "sell," "profit," and "portfolio balance," which can cause the skill to activate for loosely related conversations. In a skill that can execute real crypto trades, unintended invocation materially increases the chance of unauthorized or accidental financial actions, especially in autonomous or tool-enabled environments.

Missing User Warnings

High

• Confidence: 94% confidence
• Finding: The skill promotes autonomous trading of real assets but does not present a prominent upfront warning that actions can move funds, incur irreversible losses, and execute on-chain transactions. Although the document includes risk-management guidance, that is not a substitute for explicit user consent and clear warnings about financial risk, execution risk, and the consequences of automation.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: This documentation repeatedly presents natural-language commands for creating, modifying, and canceling automated trading actions that can directly affect real funds, but it does not clearly warn users that these are live financial operations. In the context of an autonomous Base trading skill, that omission increases the risk that a user or downstream agent treats examples as harmless guidance or simulation, leading to unintended orders, cancellations of protections, or real losses.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The document gives concrete trading advice such as taking 'more aggressive, larger positions' during perceived bull conditions without any balancing warning about volatility, losses, or suitability. In an autonomous trading skill, this can encourage unsafe risk-taking by users or downstream agents and may contribute to significant financial harm if the guidance is followed mechanically.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: This file provides trading research commands and indicator interpretations in a way that can directly influence user financial decisions, but it omits a clear warning that the outputs are informational only and may be inaccurate, delayed, or unsuitable for investment decisions. In the context of an autonomous crypto trading skill, that omission increases the chance users treat generated analysis as actionable advice and incur losses in a highly volatile market.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The pre-trade checklist and workflow present generated analysis as a procedural step before trading, without warning users to validate results or consider risks such as scams, thin liquidity, manipulation, and rapid price swings. Because this skill is specifically designed for crypto trading on Base, the operational framing makes it more likely that users will convert imperfect AI analysis into immediate trades, amplifying financial harm.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: This file gives concrete trading, position-sizing, stop-loss, and portfolio-allocation guidance without any disclaimer that the material is educational, may cause financial loss, and is not personalized financial advice. In the context of an autonomous crypto trading skill, users may treat these recommendations as actionable and safe defaults, increasing the risk of unsuitable trades and monetary harm.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The document gives concrete buy/sell, DCA, sniping, and momentum-trading instructions, including examples that encourage execution, but it does not include a clear upfront warning about the substantial risk of financial loss. In the context of an autonomous crypto trading skill, this omission is more dangerous because users may treat the content as actionable guidance and expose funds to highly volatile or scam-prone assets.

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: These example prompts are generic enough to trigger the trading skill from normal analytical requests, especially phrases like 'What's the price,' 'Do technical analysis,' and 'What tokens are trending on Base?'. In an autonomous trading skill, vague prompt examples can blur the boundary between information gathering and execution, increasing the chance of unintended activation or overbroad invocation.

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.