ClawHub
Back to skill

Security audit

Osh Contract Review

Security checks for vulnerabilities and agentic risk

Overview

This is a plain-language contract review skill with no executable code or hidden system access, though users should remember it is not a substitute for a lawyer.

Reasonable to install for plain-language contract walkthroughs. Do not rely on it for legal advice, active disputes, jurisdiction-specific interpretation, or high-value signing decisions; ask for your preferred language if the contract is not in English.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill description is broadly framed as helping someone 'review, understand, or assess a contract in plain language' across several contract types, without clearly limiting when it should or should not be invoked. That can cause over-triggering in adjacent legal or high-stakes scenarios, leading the agent to apply a generic contract-review workflow where specialized legal escalation, privacy handling, or refusal logic would be more appropriate.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The rule 'If the contract is in a language other than English, respond in that same language' overrides user preference and can force the assistant into a language the user may not read well, especially when the contract language differs from the user's preferred language. In a legal-context skill, this increases the chance of misunderstanding important terms, hiding uncertainty, and producing unsafe guidance because the user did not opt into that output language.

Static analysis

No suspicious patterns detected.