Security audit

Committee Prediction Market Skills

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed guide for BSC mainnet prediction-market contract calls, with real financial risk that users should confirm before use.

Before installing or using this skill, treat all write calls as real BSC mainnet transactions: confirm contract addresses, token approvals, amounts, deadlines, gas, slippage/min outputs, and whether funds may be locked or lost before allowing an agent to submit anything.

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
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: This skill gives detailed instructions for interacting with live BSC mainnet prediction-market contracts, including staking, trading, dispute bonds, and redemption, but does not prominently warn users that transactions are irreversible and may cause real financial loss. In this context, omission of financial-risk and on-chain finality warnings increases the chance that an agent or user will execute costly actions without informed consent, especially because the document reads like an operational runbook rather than a cautionary guide.

Natural-Language Policy Violations

Low

Confidence: 78% confidence
Finding: The manifest sets `language: en`, which imposes a specific language choice in natural-language metadata. The file does not indicate that the user can choose another language or that English is required for a documented region-specific or compliance reason.