Trust Assessment
spec-driven-eval 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 Dynamic Script Execution of Untrusted Inputs.
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 August 1, 2026 (commit 529b8103). 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 Dynamic Script Execution of Untrusted Inputs The skill instructs the agent to compute scores by executing a shell command (e.g., `node -e` or `python3 -c`) using a script constructed from the evaluation data (such as story names, priorities, and weights). If the PRD or spec being evaluated contains malicious payloads in story names or other fields, and the agent dynamically interpolates these into the shell command without strict sanitization, it can lead to arbitrary command execution on the host system. Avoid instructing the agent to execute shell commands or dynamic scripts with untrusted inputs. Instead, the agent should perform the calculations using its internal reasoning/math capabilities, or use a dedicated, sandboxed math tool/library that does not evaluate arbitrary code. | LLM | SKILL.md:70 |
Scan History
Embed Code
[](https://skillshield.io/report/9e556c1bf723ce25)
Powered by SkillShield