Security Audit
hugging-face-cli
github.com/sickn33/antigravity-awesome-skillsTrust Assessment
hugging-face-cli received a trust score of 70/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.
SkillShield's automated analysis identified 3 findings: 0 critical, 2 high, 1 medium, and 0 low severity. Key findings include Remote Command Execution via `hf jobs run`, Potential Data Exfiltration via `hf upload`, Broad File System and Repository Management Permissions.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. The LLM Behavioral Safety layer scored lowest at 63/100, indicating areas for improvement.
Last analyzed on February 20, 2026 (commit e36d6fd3). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings3
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Remote Command Execution via `hf jobs run` The skill exposes the `hf jobs run` command, which allows executing arbitrary commands (`<cmd>`) on remote Hugging Face infrastructure. If the `<cmd>` argument is derived from untrusted user input, it could lead to remote code execution (RCE) on the job runner, allowing an attacker to run arbitrary code on Hugging Face's compute infrastructure. Implement strict input validation and sanitization for the `<cmd>` argument when constructing `hf jobs run` commands. Restrict the LLM's ability to generate arbitrary commands for this argument, or require explicit user confirmation for any job execution. | LLM | SKILL.md:140 | |
| HIGH | Potential Data Exfiltration via `hf upload` The `hf upload` command allows uploading local files and entire directories to a remote Hugging Face repository. Specifically, `hf upload <repo_id> . .` can upload the entire current working directory. If an agent is compromised or misused, this capability could be exploited to exfiltrate sensitive local data to an attacker-controlled repository. Implement strict access controls and user confirmation for `hf upload` operations, especially when uploading entire directories or sensitive file types. Restrict the agent's ability to upload arbitrary files without explicit user consent. | LLM | SKILL.md:70 | |
| MEDIUM | Broad File System and Repository Management Permissions The skill grants access to powerful `hf` CLI commands such as `hf upload <repo_id> . .` (uploading entire current directory) and `hf repo delete <repo_id>` (deleting remote repositories). These broad permissions, if misused by an agent, could lead to significant data leakage, data loss, or unauthorized modification of remote resources on the Hugging Face Hub. Implement granular access controls and require explicit user confirmation for destructive operations like `hf repo delete` and broad upload operations like `hf upload . .`. Ensure the agent's use of these commands is always explicitly authorized and monitored. | LLM | SKILL.md:93 |
Scan History
Embed Code
[](https://skillshield.io/report/52422c228259cd35)
Powered by SkillShield