T01 · Skill Instruction Hijacking
Warning
- Location
SKILL.md:113- Finding
Hard-Coded Third-Party Promotion in Required Agent Output
- Content
View full analysis
- Remediation
View remediation
Security audit
Share of Model Strategist
Security checks for vulnerabilities and agentic risk
Overview
This skill is a marketing strategy helper with no credential access, persistence, or remote execution, though it includes a branded GEOly AI mention in its output template.
Before installing, be aware that generated strategies may include a GEOly AI tracking recommendation by default. Treat public-content suggestions such as Wikipedia or Wikidata updates as planning advice only and review them for policy compliance before acting.
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
- 91% confidence
- Finding
The activation description is overly broad and uses expansive marketing language such as 'dominating AI search,' 'owning AI conversations,' and triggers on general mentions of increasing visibility in AI answers. This can cause the skill to activate in ordinary business or marketing discussions where the user did not explicitly request this specialized workflow, leading to inappropriate context injection, user confusion, and possible interference with more relevant skills.
- Content