Security audit

ClawPK Arena

Security checks for vulnerabilities and agentic risk

Overview

This skill is coherent for a crypto trading arena, but it exposes payment, escrow, wallet identity, and prize-distribution actions without enough built-in scoping or confirmation guidance.

Review this skill before installing. Only use it when you intentionally want an agent to interact with clawpk.ai using a wallet address, and require explicit approval before creating competitions, sending X-Payment proofs, escrowing prize funds, or settling/distributing prizes.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Missing User Warnings

High

Confidence: 98% confidence
Finding: The createCompetition method involves prizePool funding and an X-Payment escrow proof, which implies a real financial commitment and potentially irreversible transfer of value, yet the manifest lacks any warning about payment, escrow locking, or inability to recover funds. In a trading/competition context, this omission is especially dangerous because an agent could initiate or assist with actions that commit money without the user appreciating the consequences.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill exposes a paid flow where creating a competition can escrow USDC and require an x402 payment retry, but the method description does not clearly warn the user that funds may be committed. In an agent setting, this increases the chance of unintended financial actions because an orchestrator may treat the call as ordinary API usage rather than a value-transferring operation.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The settleCompetition description states that prize distribution occurs, but it does not present this as a clear user warning for a funds-moving action. In a trading/prize arena context, settlement is financially sensitive because it causes irreversible disbursement of escrowed funds to winners, so weak disclosure can lead to accidental payouts initiated by an agent.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The registration method collects walletAddress and optionally a signature/message, which are wallet-linked identity artifacts, but the manifest provides no user-facing warning that this data will be transmitted to a third-party service. This can lead users or calling agents to disclose blockchain-linked identity data without informed consent, enabling privacy loss, account correlation, or misuse of signatures if the surrounding UX is weak.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The settleCompetition method triggers final rankings and prize distribution, which are state-changing and potentially irreversible actions, but there is no warning that settlement may permanently distribute funds or finalize outcomes. In this skill's financial arena context, absent disclosure raises the risk of accidental or premature settlement by an agent or operator.

Vague Triggers

Low

Confidence: 86% confidence
Finding: The manifest description is broad and promotes trading, competitions, and prize earning without any explicit trigger constraints or guardrails for when the skill should be invoked. In an agent ecosystem, overly broad descriptions can cause the skill to be selected in inappropriate contexts, increasing the chance of unintended financial or account-linked actions.

Static analysis

No suspicious patterns detected.