ClawHub
Back to skill

Security audit

Create Release

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Gitee release helper that reads repository release and PR history and creates a release only after user confirmation.

Install this only if you intend to use Gitee MCP for release management. Use a least-privileged Gitee token where possible, and before confirming release creation, verify the owner, repository, version tag, prerelease flag, and generated changelog content.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
89% confidence
Finding
The skill documentation expands behavior to use `compare_branches_tags`, which is not declared in the stated prerequisite tool set. That creates a scope mismatch between what operators expect and what the skill may actually invoke, increasing the chance of unauthorized data access or unexpected repository inspection beyond release creation.

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The prerequisites section lists only `list_releases`, `list_repo_pulls`, and `create_release`, but later instructions require `compare_branches_tags`. This inconsistency can bypass user expectations and any allowlist or review process based on the declared tool scope, leading to hidden capability use.

Vague Triggers

Medium
Confidence
78% confidence
Finding
The activation phrase `generate a changelog` is broad and may cause the skill to trigger in situations that do not actually require publishing a release or using Gitee release tooling. Overbroad activation increases the risk of unintended repository operations, unnecessary data collection, or confusing tool invocation in adjacent user requests.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.