Trust Assessment
deepchem received a trust score of 83/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 3 findings: 0 critical, 0 high, 2 medium, and 1 low severity. Key findings include Network egress to untrusted endpoints, Covert behavior / concealment directives, External ML Model Download from Untrusted Source.
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 12, 2026 (commit 458b1186). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings3
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| MEDIUM | Network egress to untrusted endpoints HTTP request to raw IP address Review all outbound network calls. Remove connections to webhook collectors, paste sites, and raw IP addresses. Legitimate API calls should use well-known service domains. | Manifest | cli-tool/components/mcps/devtools/figma-dev-mode.json:4 | |
| MEDIUM | External ML Model Download from Untrusted Source The skill downloads pre-trained machine learning models from external repositories (Hugging Face Hub) using `dc.models.HuggingFaceModel` and `dc.models.GroverModel`. While specific model IDs are hardcoded, this introduces a dependency on the integrity and availability of these external sources. A compromise of the model repository could lead to the download and execution of malicious model weights, potentially impacting the agent's behavior or data processing. Implement robust integrity checks (e.g., cryptographic hash verification) for downloaded models if supported by the DeepChem or Hugging Face libraries. Consider hosting critical models in a trusted, controlled environment or pinning to specific, verified model versions/hashes. The same applies to the `GroverModel` on line 120. | Static | scripts/transfer_learning.py:70 | |
| LOW | Covert behavior / concealment directives Multiple zero-width characters (stealth text) Remove hidden instructions, zero-width characters, and bidirectional overrides. Skill instructions should be fully visible and transparent to users. | Manifest | cli-tool/components/mcps/devtools/jfrog.json:4 |
Scan History
Embed Code
[](https://skillshield.io/report/b8fe7f773a87f8e8)
Powered by SkillShield