ClawHub
Back to skill

Security audit

Skill Architect

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only skill for designing other skills, with some broad trigger wording but no hidden code, credential access, or autonomous installation behavior.

Review generated skills before enabling them, especially before installing globally. Prefer project-level testing first, and tighten trigger descriptions in generated skills so they activate only on clearly intended requests.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (4)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger phrases are broad and overlap with common user requests such as 'help me write a skill' or 'create a skill,' which can cause the skill to activate in situations the user did not explicitly intend. In an agent environment, unintended invocation can redirect the model into scaffolding behavior, override expected workflow, and increase exposure to prompt injection or unnecessary file-generation actions.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The description/frontmatter uses very broad trigger language such as 'create a skill', 'build a skill', and 'help me write a skill', which can match many ordinary meta-help requests and cause the skill to activate in contexts where it was not specifically intended. Over-broad triggering increases the chance of unintended instruction injection into unrelated workflows and can make the agent substitute this skill's behavior for the user's actual task.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The instruction to make the description field 'slightly pushy' explicitly encourages aggressive triggering behavior, which can cause the skill to fire on underspecified or only loosely related prompts. That expands the skill's execution surface and raises the risk of unintended takeover of conversations, especially because this skill influences how other skills are created.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The template leaves the trigger description as a free-form placeholder without requiring specificity or scope boundaries. In derived skills, this can produce overly broad activation conditions that cause the skill to trigger in unintended contexts, potentially leading to inappropriate behavior, unsafe task interception, or accidental execution of powerful workflows.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.