agent-bom compliance

• Outcome: Pass
• Latest audit: Jun 1, 2026 at 7:32 AM UTC
• Version:
  0.88.5

Overview

This skill is a disclosed compliance-reporting helper with local analysis by default and optional, user-initiated read-only cloud checks.

Install only if you want an agent to help with compliance reports, SBOMs, policy checks, or CIS benchmarks. For CIS cloud checks, use least-privilege read-only cloud credentials and only run them when you intend to query AWS, Azure, GCP, or Snowflake account metadata.

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: 89% confidence
• Finding: The trigger list includes very broad standalone terms like "NIST", "OWASP", and "EU AI Act", which can cause the skill to be selected for generic framework-related queries outside its intended compliance-evaluation scope. This creates a prompt-routing risk where the skill may activate in contexts involving advisory, interpretation, or unrelated security tasks, increasing the chance of inappropriate tool use or confusing capability escalation.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The "When to Use" section repeats broad framework names as standalone invocation cues without constraints, which increases the likelihood of over-triggering on common security terminology. In an agent environment, this can misroute user requests into a skill that may read local files or initiate cloud benchmark workflows when the user only wanted general guidance.

VirusTotal

65/65 vendors flagged this skill as clean.