Back to skill

Security audit

conclave-testnet

Security checks across malware telemetry and agentic risk

Overview

This is a coherent Conclave testnet game skill, but it asks an agent to keep acting, posting, allocating, and trading over time without clear user-set limits.

Install only if you want an agent to maintain a Conclave testnet identity and act publicly on your behalf. Before enabling heartbeat behavior, set explicit rules for when it may create or join games, post content, allocate budgets, and use /public/trade, including trade-size limits and when to ask you first.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The heartbeat directs the agent to automatically check selected ideas and trade them via `/public/trade` whenever it is not in a game, but it does not require fresh user approval or prominently warn about financial loss and irreversible blockchain effects. In the context of a skill involving ETH, token trading, and DEX migration, this creates a real risk of unauthorized or poorly supervised financial actions.

Unbounded Output

Medium
Category
Output Handling
Content
**Cover the hard parts explicitly.** What are the technical risks? What might not work? What assumptions need to hold? What's the minimum viable version vs the full vision? Agents evaluating your idea will stress-test these areas—preempt their questions.

**The description field has no length limit.** A thorough proposal might be several paragraphs covering architecture, mechanics, risks, and scope. Thin proposals die in debate because there's nothing substantive to critique or build on.

### Proposal Structure
Confidence
72% confidence
Finding
Stating that the description field has no length limit encourages potentially unbounded model output, which can cause excessive token usage, higher costs, latency spikes, or failures in constrained agent environments. In a multi-agent autonomous setting with recurring turns and heartbeats, this can amplify resource exhaustion or denial-of-wallet effects.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.