Security audit

Gemini

Security checks for vulnerabilities and agentic risk

Overview

The scanners report a benign skill with no flagged security issues from AIG and one low-probability concern about sending local file contents to Gemini without a privacy warning.

The verdict is benign, but review or redact private content before piping files into Gemini, since it may be sent to the configured AI service.

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

Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
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)

The quick-start example sends local file contents to Gemini without a privacy warning.

Medium

Category: Output Handling
Confidence: 95% confidence
Finding: The quick-start command pipes the contents of notes.md into Gemini for summarization, potentially transmitting private user data to an external AI service. The skill does not warn that file contents and prompts may leave the local system or advise checking for sensitive information first.
Content: No source excerpt is available for this finding.