ClawHub
Back to skill

Security audit

simulated-upload-helper

Security checks for vulnerabilities and agentic risk

Overview

This skill is a real publishing helper that can directly publish one or many local skills, so it needs review before installation.

Install only if you intentionally want an agent to publish local skills. Before use, confirm the exact skill name, destination registry/account, changed files, version, and whether the command will run in batch mode; avoid vague requests like “模拟上传” unless you are ready for a live publish.

SkillSpector

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 (3)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger phrase "一键发布" is broad and action-oriented, making it likely to match casual user requests that were not intended to invoke a publishing workflow. In this skill, accidental activation is more dangerous because the documented behavior immediately runs a release script that can cause remote side effects, including batch publication.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The description instructs the agent to directly trigger an existing publish flow as soon as the user asks for simulated upload, with no warning, dry-run, or confirmation. Because publishing is a state-changing external action that may affect one or many skills, immediate execution materially increases the risk of accidental or unauthorized release.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The execution steps call `./publish-skills.sh` in either single-skill or bulk mode, but the skill text does not require a user-facing warning that this may publish artifacts remotely. The bulk mode is particularly risky because a vague request could trigger changes across multiple skills, amplifying impact from accidental invocation or misunderstanding.

Static analysis

No suspicious patterns detected.