Nm Gauntlet Graph Build

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent local code-graph builder that writes a project-local SQLite index, with no evidence of hidden exfiltration or destructive behavior.

Install only if you want this skill to index the selected repository locally. Verify that CLAUDE_PLUGIN_ROOT points to a trusted Gauntlet/Claude Code plugin, expect .gauntlet/graph.db and possibly .gauntlet/.gitignore to be created or updated, and delete .gauntlet if you do not want the local graph retained.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill instructs the agent to build or update a SQLite knowledge graph in the current repository, which causes local filesystem writes, but the user-facing description does not clearly warn about that side effect. This can lead to unintended repository modification, dirty working trees, or confusion in sensitive environments where write operations should be explicit and consented to.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill states that it auto-creates `.gauntlet/.gitignore`, but this side effect is only disclosed in the storage section rather than as a clear operational warning. Silent creation of ignore rules can surprise users, alter repository state, and in some workflows hide generated artifacts from version control review without explicit consent.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal