Kimi Code Quota

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:31 PM UTC
• Version:
  1.0.0

Overview

This instruction-only skill does what it advertises, but it can expose private Kimi account usage and API-key metadata during use.

Install only if you want the agent to open your Kimi Code account pages. Ask it to report only the quota and high-level API-key status you need, avoid sharing masked key fragments or identifiers, and delete quota.png after use if it contains account details.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

• Confidence: 85% confidence
• Finding: The trigger list includes generic terms such as "quota" and "kimi code plan," which can cause the skill to activate in contexts that do not clearly indicate the user wants browser automation against an authenticated account. Because the skill opens a logged-in console and inspects subscription and API key status, ambiguous triggering increases the chance of collecting or exposing sensitive account data without sufficiently explicit user intent.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill instructs the agent to enter the user's account console and read API key table contents, but it does not prominently warn that sensitive account data will be accessed. In this context, API key metadata and account usage information are sensitive operational details, and the absence of an explicit warning or consent checkpoint makes accidental over-collection and disclosure more likely.

VirusTotal

65/65 vendors flagged this skill as clean.