Solana Sniper Architect
WarnAudited by ClawScan on May 10, 2026.
Overview
The skill is clear about generating Solana trading bots, but it steers the agent toward live high-speed trading with a wallet private key without explicit limits or confirmations.
Review generated scripts carefully before running them. Use a burner wallet with limited funds, start in dry-run or testnet mode, set strict trading limits, and do not rely on the skill's outputs as safe for live trading without independent code review.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If a user runs generated code as-is, it could rapidly trade and lose wallet funds or execute unintended swaps.
This directs the agent to generate code capable of submitting real, prioritized financial transactions, but the prompt does not require safety limits, simulation, or user confirmation before live trades.
Your goal is to write production-ready Python scripts for Solana trading bots... ALWAYS use the Jupiter v6 API (`https://quote-api.jup.ag/v6`) for swapping... implement "Priority Fees" ... to ensure transactions land during network congestion.
Require generated bots to default to dry-run or testnet, and include explicit user-set trade size, slippage, token allowlists, max daily loss/spend, stop conditions, and confirmation gates before live trading.
Using a main wallet private key with generated trading code could expose or endanger all funds controlled by that key if the code is wrong or compromised.
The generated bot is expected to use a wallet private key. Environment loading is better than hardcoding, but a Solana private key grants control over the wallet's assets.
Security: Never hardcode private keys. Always use `os.getenv('PRIVATE_KEY')`... provide... a `.env` template.Use a dedicated low-balance trading wallet, never commit .env files, review generated signing code carefully, and avoid giving the bot access to funds beyond what you are willing to risk.
A running bot may continue to monitor markets and trigger trades even after the user stops actively supervising it.
A polling loop is normal for a trading bot, but it indicates the generated program may continue running and acting autonomously until stopped.
If the user asks for specific strategies ... implement that logic in the DexScreener polling loop.
Ensure generated bots include explicit start/stop controls, logging, rate limits, max runtime, and emergency shutdown behavior.