Uniswap Seek Protocol Fees
WarnAudited by ClawScan on May 10, 2026.
Overview
The skill is coherent for Uniswap fee analysis, but it can direct an agent toward irreversible wallet-funded transactions and relies on unreviewed subagent components.
Only install this if you are comfortable with an agent analyzing and potentially initiating Uniswap transactions. Keep it in preview mode unless you intentionally want execution, use a limited-purpose wallet, and verify the referenced Uniswap MCP tools and subagents before allowing any burn, claim, or swap.
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.
A user could unintentionally authorize a high-value blockchain transaction if they do not understand that claiming fees requires burning 4,000 UNI.
The skill can treat user wording such as 'claim fees' as permission to execute an irreversible on-chain burn-and-claim, rather than only presenting a preview.
auto-execute | No | false | "execute the burn", "claim fees" implies true; "check", "preview" implies false
Keep preview-only as the default, require an explicit transaction-level confirmation, and show recipient, UNI spend, gas, calldata, slippage, and expected proceeds before execution.
The agent may operate with wallet authority capable of spending significant funds.
The workflow depends on a connected or agent-controlled wallet that may hold enough UNI to perform the burn, but the registry metadata declares no primary credential or required environment variables.
get_agent_balance — does agent have enough UNI? ... recipient | No | connected wallet | Explicit address if provided, otherwise agent's wallet
Use a dedicated wallet with limited funds, declare the wallet/account requirements clearly, and enforce spend limits and user confirmations outside the skill text.
Important execution and safety behavior may depend on components not included in the reviewed artifacts.
The skill relies on named subagents for profitability and safety decisions, but the supplied package contains only README.md and SKILL.md, so those subagent definitions are not reviewable here.
allowed-tools: - Task(subagent_type:protocol-fee-seeker) - Task(subagent_type:safety-guardian)
Provide and pin the referenced subagent definitions or require users to verify their provenance and tool permissions before enabling execution.