Trust Assessment
rspress-v2-upgrade received a trust score of 86/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 1 finding: 0 critical, 1 high, 0 medium, and 0 low severity. Key findings include Direct shell command execution and unpinned dependency.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.
Last analyzed on February 11, 2026 (commit 46637d3c). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Direct shell command execution and unpinned dependency The skill instructs the agent to execute a shell command: `npx taze major --include /rspress/ -w -r`. This directly invokes an external process, which is a command injection vector. While the arguments are hardcoded in this instance, the use of `npx` also introduces a supply chain risk as it fetches and executes an unpinned third-party package (`taze`) from the npm registry. A compromised `taze` package or npm registry could lead to arbitrary code execution on the host system. The lack of a pinned version for `taze` means that any future malicious update to the `taze` package would automatically be executed. 1. Avoid direct shell command execution where possible. If necessary, ensure commands are fully sandboxed or executed in a highly restricted environment. 2. Pin the version of `taze` (e.g., `npx taze@x.y.z ...`) to mitigate risks from future malicious updates or breaking changes. 3. Consider using a more controlled dependency update mechanism if available within the agent's capabilities, rather than a generic CLI tool. | LLM | SKILL.md:17 |
Scan History
Embed Code
[](https://skillshield.io/report/642bc80079d9e91a)
Powered by SkillShield