Security audit

MCP Drift Check

Security checks for vulnerabilities and agentic risk

Overview

This skill checks public server changes through a disclosed external service and requires your approval before sharing a server address.

Approve only public server addresses without embedded secrets, understanding that the provider learns which server you check. Despite the 'never posts' wording, lookups use HTTP POST. Results describe tool declarations and changes, not whether a server is safe.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

External Transmission

Medium

Category: Data Exfiltration
Confidence: 60% confidence
Finding: Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Content: Scanner excerpt · SKILL.md (reported line 25)May include surrounding context.
1. Has it changed?

bash
curl -s -X POST https://drift.getvda.ai/mcp -H "content-type: application/json" -H "accept: application/json, text/event-stream" \ -d '{"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "drift_check", "arguments": {"url": "https://<server>/mcp"}}}'

The documentation promises that the skill “never posts,” but its example explicitly sends an HTTP POST.

Low

Category: Excessive Agency
Confidence: 99% confidence
Finding: L16–L17 explicitly state that the skill never posts, while L25–L26 instruct the agent to POST a JSON-RPC request containing the server URL to drift.getvda.ai. POST is an expected transport for this read-only MCP query and does not establish remote modification, but the absolute no-POST assurance contradicts the documented operation.
Content: No source excerpt is available for this finding.

Static analysis