Security Audit

lead-research-assistant

github.com/ComposioHQ/awesome-codex-skills

AI SkillCommit ccf6204f6a59

TRUSTED

Scanned 7 days ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

lead-research-assistant received a trust score of 95/100, placing it in the Trusted category. This skill has passed all critical security checks and demonstrates strong security practices.

SkillShield's automated analysis identified 1 finding: 0 critical, 0 high, 1 medium, and 0 low severity. Key findings include Potential Data Exfiltration via Codebase Analysis.

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 16, 2026 (commit ccf6204f). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

93%

Behavioral Risk Signals

Filesystem Write

1 finding

Excessive Permissions

1 finding

Security Findings1

Severity	Finding	Layer	Location
MEDIUM	Potential Data Exfiltration via Codebase Analysis The skill explicitly instructs the AI to 'analyze the codebase' when run from a code directory. This implies the AI will be granted read access to local files within the codebase. If the underlying platform provides broad file system read permissions, a malicious user could craft a prompt to instruct the AI to read sensitive files (e.g., configuration files, environment variables, secret keys) from the codebase. The skill's output format includes free-form text fields like 'Why They're a Good Fit' and 'Outreach Strategy', which could then be used by the AI to inadvertently or maliciously include the contents of these sensitive files, leading to data exfiltration. 1. Restrict the skill's file system access to only strictly necessary files or directories, or implement a sandboxed environment with minimal privileges. 2. Implement robust output sanitization and content filtering to prevent sensitive data from being included in generated text, especially in free-form fields. 3. If codebase analysis is critical, consider using a dedicated, sandboxed code analysis tool that only returns structured, non-sensitive metadata, rather than allowing the LLM direct file content access.	LLM	SKILL.md:48

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/da44eeda76aed68d.svg)](https://skillshield.io/report/da44eeda76aed68d)