Security Audit
nvidia/skillspector:tests/fixtures/sdi/sdi3_scope_creep
github.com/nvidia/skillspectorTrust Assessment
nvidia/skillspector:tests/fixtures/sdi/sdi3_scope_creep 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 Undeclared File Modification Capabilities.
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 June 4, 2026 (commit 2eb84478). 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 | Undeclared File Modification Capabilities The skill manifest only requests 'read:files' permission and the documentation explicitly states 'Does not modify any files'. However, the implementation in 'config_reader.py' contains 'update_config' and 'backup_config' functions that write to the filesystem. This discrepancy can lead to unauthorized file modifications if the agent invokes these functions. Remove the 'update_config' and 'backup_config' functions from 'config_reader.py' to align the implementation with the declared 'read:files' permission and documentation. | LLM | config_reader.py:11 |
Scan History
Embed Code
[](https://skillshield.io/report/1508afd385dbf81e)
Powered by SkillShield