Instagram Publisher

Security checks across malware telemetry and agentic risk

Overview

This Instagram publishing skill does what it advertises, but its security section understates that it can upload media and create or schedule public posts through a credential-bearing connector.

Install only if you trust Boring with publish access to the connected Instagram account. Treat the MCP URL like a password, connect only accounts you intend the agent to use, upload only media meant for public posting, and require the agent to show final account, media, caption, and schedule details before publishing or canceling anything.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The skill’s data-handling section is internally inconsistent: it claims only analytics metrics are retrieved and that no content is uploaded or modified, while later steps explicitly instruct uploading media and publishing or scheduling Instagram posts. This misleading disclosure can cause users or downstream agents to underestimate the write capabilities and sensitivity of the embedded MCP token, increasing the chance of unsafe use or overbroad trust.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal