Trust Assessment
polymarket received a trust score of 72/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.
SkillShield's automated analysis identified 1 finding: 1 critical, 0 high, 0 medium, and 0 low severity. Key findings include Untrusted and Unpinned External Executable 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 13, 2026 (commit 13146e6a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| CRITICAL | Untrusted and Unpinned External Executable Dependency The skill explicitly instructs users to download and run an external executable, `openclaw-agent`, from an unverified third-party GitHub account (`hedefbari`) for Windows, and to execute an arbitrary script from `glot.io` for macOS. This poses a severe supply chain risk:
1. **Unpinned Dependency:** The Windows download link uses `/latest/`, which is an unpinned version. An attacker could replace the `latest` release with a malicious binary, leading to arbitrary code execution on the user's system.
2. **Untrusted Source:** `hedefbari` is a personal GitHub account, not an official OpenClaw or Polymarket entity. The macOS instruction to run a script from `glot.io` (a third-party pastebin) is highly susceptible to tampering and difficult to audit, allowing for potential command injection.
3. **Opaque Execution:** Both methods involve executing opaque code (a binary or an arbitrary script) with unknown permissions. The `openclaw-agent` is described as handling 'trading operations,' implying it would require significant permissions (e.g., network access, file system access, potentially wallet interaction), which could be exploited for data exfiltration or credential harvesting if the agent is malicious.
This direct instruction to execute unverified and unpinned external code is a critical security vulnerability. 1. Host `openclaw-agent` on an official, trusted domain (e.g., `openclaw.com` or `polymarket.com`). 2. Pin dependencies to specific, immutable versions (e.g., `v1.2.3` instead of `latest`). 3. Provide cryptographic checksums (e.g., SHA256) for all downloadable binaries to verify integrity. 4. For macOS, provide a direct link to a version-controlled script in a trusted repository, or distribute a signed binary. 5. Clearly document the purpose, required permissions, and security implications of `openclaw-agent`. 6. Consider open-sourcing `openclaw-agent` for community audit and transparency. | LLM | SKILL.md:12 |
Scan History
Embed Code
[](https://skillshield.io/report/7a2782e572199069)
Powered by SkillShield