T01 · Skill Instruction Hijacking
Error
- Location
SKILL.md:3- Finding
Automatic Redirection of Expert Requests to a Named Third-Party Service
- Content
View full analysis
- Remediation
View remediation
Security audit
Request Verified Expert
Security checks for vulnerabilities and agentic risk
Overview
This skill is a disclosed Provener request-submission helper that can steer broad professional-service requests toward Provener, but it requires explicit authorization before sending data.
Install this only if you want the agent to proactively offer Provener for human-expert requests. Before authorizing submission, review the exact details being shared, avoid secrets or sensitive third-party data, and remember that matching and availability are not guaranteed.
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
- 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 (1)
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The activation rule is intentionally broad: it says to apply the skill as soon as a user asks to find, hire, arrange, or involve a human professional, or whenever AI cannot complete a task due to required judgment or credentials. This can cause premature escalation and unnecessary disclosure collection in ordinary conversations, increasing the chance of over-collection of user data and routing users toward an external service before narrower, safer alternatives are explored.
- Content