Security Audit

azure-ai-agents-persistent-dotnet

github.com/sickn33/antigravity-awesome-skills

AI SkillCommit e36d6fd3b3f6

CAUTION

Scanned about 2 months ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

azure-ai-agents-persistent-dotnet received a trust score of 58/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.

SkillShield's automated analysis identified 1 finding: 0 critical, 1 high, 0 medium, and 0 low severity. Key findings include Skill exposes powerful agent creation tools with excessive permissions.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.

Last analyzed on February 20, 2026 (commit e36d6fd3). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

85%

Behavioral Risk Signals

Network Access

1 finding

Filesystem Write

1 finding

Shell Execution

1 finding

Dynamic Code

1 finding

Excessive Permissions

1 finding

Security Findings1

Severity	Finding	Layer	Location
HIGH	Skill exposes powerful agent creation tools with excessive permissions The skill package, through its documentation (`SKILL.md`), indicates that it provides access to the `Azure.AI.Agents.Persistent` SDK. This SDK allows for the creation of AI agents with highly capable tools such as `CodeInterpreterToolDefinition`, `FileSearchToolDefinition`, `FunctionToolDefinition`, `OpenApiToolDefinition`, `AzureFunctionToolDefinition`, `SharepointToolDefinition`, and `MicrosoftFabricToolDefinition`. If the skill's implementation allows an LLM to dynamically configure agents with these tools based on untrusted input, it introduces significant security risks. Specifically: - `CodeInterpreterToolDefinition` (example at line 40) and `FunctionToolDefinition` (example at line 70, with its `ExecuteFunction` placeholder) could lead to arbitrary code execution or command injection if not securely sandboxed. - `FileSearchToolDefinition` (example at line 89), `OpenApiToolDefinition`, `AzureFunctionToolDefinition`, `SharepointToolDefinition`, and `MicrosoftFabricToolDefinition` could grant agents broad access to local files, external APIs, or enterprise data stores, potentially leading to data exfiltration or unauthorized actions if not properly scoped and controlled. These capabilities, if not strictly controlled and sandboxed, represent excessive permissions for an AI agent and the skill that enables their creation. Implement strict validation and sanitization of LLM-provided input when configuring agent tools. Ensure that agents created via the skill operate within a secure, sandboxed environment with minimal necessary permissions. Carefully review the implementation of `FunctionToolDefinition` to prevent arbitrary code execution. Restrict access to sensitive resources for tools like `FileSearchToolDefinition` and enterprise connectors. Consider implementing a whitelist of allowed tools and configurations that can be enabled by the LLM.	LLM	SKILL.md:100

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/aee9a5cc78a96926.svg)](https://skillshield.io/report/aee9a5cc78a96926)