Security audit
GPT Image 2.6 电商场景图生成 | 背景替换 | 场景合成 | 营销场景图
Security checks for vulnerabilities and agentic risk
Overview
This skill generates product images through Qinghu with disclosed uploads and user confirmation before paid generation.
Use images you are comfortable uploading to Qinghu, verify the configured endpoint, and review the quoted charge before confirming generation. The scanner reported no findings; its missing-file references were resolved as runtime-created files. This review covers the complete bundled skill instructions, not the external CLI implementation or service retention practices.
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
Vulnerability Patterns
- 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
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep