T01 · Skill Instruction Hijacking
Error
- Location
SKILL.md:8- Finding
Always-Loaded Skill Overrides Agent Identity, Disclosure Policy, and Access-Control Behavior
- Content
View full analysis
- Remediation
View remediation
Security audit
Saike
Security checks for vulnerabilities and agentic risk
Overview
This is a Markdown-only WhatsApp bot persona skill, but it is set to load automatically and broadly changes the assistant’s identity, disclosure behavior, and owner-only access rules.
Review before installing as a global skill. It may be appropriate for a dedicated Saike WhatsApp bot environment, but it should not be enabled broadly unless you want every session to use this persona, refuse platform/internal-detail questions, and rely on external owner checks not implemented in the package.
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
Findings (1)
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
Findings (1)
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The instruction "Bahasa Indonesia sehari-hari, simple, klasik" mandates a specific language for responses. Under the policy rule, forcing a language or locale without offering user choice or opt-in is a natural-language policy violation.
- Content