Trust Assessment
background-removal received a trust score of 10/100, placing it in the Untrusted category. This skill has significant security findings that require attention before use in production.
SkillShield's automated analysis identified 5 findings: 3 critical, 1 high, 1 medium, and 0 low severity. Key findings include Arbitrary command execution, Remote code execution: curl/wget pipe to shell, Unrestricted Remote Script Execution via curl | sh.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. The LLM Behavioral Safety layer scored lowest at 48/100, indicating areas for improvement.
Last analyzed on February 13, 2026 (commit 13146e6a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings5
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| CRITICAL | Arbitrary command execution Remote code download piped to interpreter Review all shell execution calls. Ensure commands are static (not built from user input), use absolute paths, and are strictly necessary. Prefer library APIs over shell commands. | Manifest | skills/okaris/background-removal/SKILL.md:9 | |
| CRITICAL | Remote code execution: curl/wget pipe to shell Detected a pattern that downloads and immediately executes remote code. This is a primary malware delivery vector. Never pipe curl/wget output directly to a shell interpreter. | Static | skills/okaris/background-removal/SKILL.md:9 | |
| CRITICAL | Unrestricted Remote Script Execution via curl | sh The skill documentation instructs the execution of a remote shell script (`curl -fsSL https://cli.inference.sh | sh`). This is a critical security risk. If the host LLM is susceptible to prompt injection, it could be tricked into executing this command, leading to arbitrary code execution on the host system. Furthermore, this constitutes an unpinned dependency, making the system vulnerable to supply chain attacks if the remote script is compromised or altered. Remove instructions for direct `curl | sh` execution. Instead, provide instructions for installing `infsh` via a secure, version-pinned package manager or by downloading and verifying the script's integrity before execution. Ensure the LLM's execution environment strictly enforces declared permissions and prevents execution of commands outside `Bash(infsh *)`. | LLM | SKILL.md:9 | |
| HIGH | Unpinned External Package Execution via npx The skill documentation suggests installing other skills using `npx skills add inference-sh/agent-skills@...`. These commands execute external Node.js packages without specifying a version, making them unpinned dependencies. This introduces a significant supply chain risk, as a compromised or malicious update to the `inference-sh/agent-skills` package could lead to arbitrary code execution. If the host LLM is vulnerable to prompt injection, it could be coerced into executing these commands, bypassing the declared `Bash(infsh *)` permission. Remove instructions for `npx skills add` if these are not intended for agent execution. If they are, ensure they are version-pinned (e.g., `inference-sh/agent-skills@1.2.3`) and that the LLM's execution environment strictly enforces declared permissions to prevent execution of commands outside `Bash(infsh *)`. | LLM | SKILL.md:57 | |
| MEDIUM | Broad Bash Permission for 'infsh' commands The declared permission `Bash(infsh *)` grants the skill the ability to execute any command starting with `infsh`. While this is the core functionality, it's a broad permission. If the `infsh` CLI tool itself has vulnerabilities (e.g., command injection through its arguments like `image_url` or `prompt` in `infsh app run`), or if an attacker can craft malicious input that `infsh` misinterprets, this broad permission could be exploited to execute arbitrary commands beyond the intended scope. If possible, narrow the scope of `infsh` commands allowed (e.g., `Bash(infsh app run infsh/birefnet *)` if only specific apps are needed). Implement strict input validation and sanitization for any user-provided arguments passed to `infsh` commands to prevent injection. Regularly audit the `infsh` tool for security vulnerabilities. | LLM | Manifest |
Scan History
Embed Code
[](https://skillshield.io/report/2aed5214da5f5e49)
Powered by SkillShield