AMG Azure Subscription Cost Analysis

Security checks across malware telemetry and agentic risk

Overview

This skill performs the Azure cost analysis it advertises, but users should treat saved cost reports and any approved parsing commands as sensitive.

Install only if you intend to connect it to your own Azure Managed Grafana MCP endpoint. Use a least-privilege Grafana service-account token, review any shell or interpreter command before approving it, and delete memory/amg-check-azure-spend/report.md when you no longer want the detailed cost report retained.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The error-handling guidance tells the agent to save oversized results to a temp file and then process them with arbitrary local interpreters or shell-capable tools such as node, python, jq, or pwsh. That expands the skill from Azure cost retrieval into unrestricted local code execution based on runtime data and environment availability, creating a clear path for unsafe command construction, data exfiltration, or execution of attacker-influenced content.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill instructs the agent to persist a full Azure cost report, including subscription names/IDs, spending by region, and service/category breakdowns, into a memory file without any user consent, minimization, retention limit, or warning that this data will be stored. Billing and subscription metadata can be sensitive operational and financial information, and persistence increases exposure to later unintended disclosure through other prompts, skills, or workspace access.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The guidance explicitly instructs writing potentially sensitive billing data to a temporary file and executing an interpreter to parse it, without any constraints on file location, retention, quoting, or data sensitivity. In the context of a cost-analysis skill, this is unnecessary and increases risk of local data leakage, unsafe processing of untrusted content, and unintended system-side effects from shell/interpreter invocation.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal