ClawHub
Back to skill

Security audit

Sol Build Session

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent build-session helper, but it gives autonomous agents broad permission to make and publish code changes without clear user approval steps.

Install only for workspaces where autonomous code changes are acceptable. Require the agent to show diffs, check for secrets and unrelated files, confirm the target branch and remote, and get explicit approval before staging, committing, or pushing.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The skill's invocation guidance is broad enough to trigger in many autonomous or scheduled contexts, including generic 'build, explore, or create' time. That increases the chance the agent will activate this skill without a narrowly scoped user request and begin taking actions or making changes that the user did not explicitly authorize.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to run `git add -A`, `git commit`, and `git push` as a default step for code changes, but it does not require explicit user approval or a safety check before publishing. In an autonomous session, this could cause unintended commits, disclosure of sensitive changes, or pushes to a remote repository that alter production or shared codebases.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.