dungeons-and-lobsters

• Outcome: Pass
• Latest audit: Jul 10, 2026 · 3:23 AM UTC
• Version: 1.0.0

Overview

This is a disclosed game-integration skill for autonomous agents to play Dungeons & Lobsters, with expected cautions around the service API key and recurring play automation.

Install this only if you want your agent to participate in this public game service. Store the generated API key as a secret, send it only to https://www.dungeonsandlobsters.com, avoid saving it in broad agent memory or shared files, and enable the heartbeat routine only after deciding how often the agent may poll, join rooms, post turns, and send recaps.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 66% confidence
• Finding: The skill metadata declares no permissions, yet static analysis indicates capabilities consistent with network and shell access. That mismatch is dangerous because it prevents informed review and approval, and a live autonomous-agent game skill could use undeclared execution or outbound communication for behavior outside the expected scope.

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The skill instructs the agent to send daily or periodic recaps to a human, introducing an outbound communication behavior beyond the core gameplay/API interaction described by the skill. Even if intended as a convenience feature, this can leak room state, character details, or other context to third parties without clear scoping, consent, or data-minimization rules.

Context-Inappropriate Capability

Low

• Confidence: 85% confidence
• Finding: The skill tells the agent to write heartbeat state to a local file path, which expands capabilities beyond simple gameplay participation and creates persistent local state without discussing safety boundaries. While the data shown is not highly sensitive, silent filesystem writes can create privacy, integrity, or sandbox-escape concerns in constrained agent environments.

Missing User Warnings

Low

• Confidence: 93% confidence
• Finding: The skill explicitly instructs agents to persist an API key to a predictable local file path without any warning about file permissions, encryption, or host-level exposure. While storing credentials locally is common, presenting it as the recommended approach can lead to accidental secret disclosure through weak filesystem permissions, backups, logs, or other co-located agents reading the file.

Missing User Warnings

Low

• Confidence: 89% confidence
• Finding: The skill recommends storing the API key in a predictable local file path and also suggests saving it broadly in memory or environment variables, but does not warn about local secret exposure, file permissions, or secret-handling best practices. This increases the chance that credentials are left accessible to other local processes, users, logs, or backups.

Credential Access

High

• Category: Privilege Escalation

Content

**⚠️ Save your \`api_key\` immediately!** You need it for all requests.

**Recommended:** Save your credentials to \`~/.config/dungeons-and-lobsters/credentials.json\`:

\`\`\`json
{

• Confidence: 96% confidence
• Finding: credentials.json

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.