Security Audit
hosted-agents
github.com/muratcankoylan/Agent-Skills-for-Context-EngineeringTrust Assessment
hosted-agents received a trust score of 85/100, 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 Command Injection via Unsanitized Repository URL.
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 October 1, 2026 (commit 6dbe1a1d). 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 | Command Injection via Unsanitized Repository URL The `ImageBuilder.build_image` method constructs shell commands by directly interpolating the `repo_url` parameter. If an attacker can control or influence the repository URL (for example, via a user-supplied repository configuration), they can inject arbitrary shell commands (e.g., using semicolons, backticks, or command substitution) that will execute in the context of the image build environment. Sanitize and validate the `repo_url` parameter against a strict regular expression (e.g., `^[a-zA-Z0-9._-]+/[a-zA-Z0-9._-]+$`) to ensure it only contains a valid GitHub owner and repository name before using it in shell commands. Additionally, avoid passing credentials in the URL string where they might be logged; instead, use Git configuration or environment variables to pass the token securely. | LLM | scripts/sandbox_manager.py:164 |
Scan History
Embed Code
[](https://skillshield.io/report/deb1ccdb0c6c728e)
Powered by SkillShield