Security Audit

context-degradation

github.com/muratcankoylan/Agent-Skills-for-Context-Engineering

AI SkillCommit 7942df36cfca

TRUSTED

Scanned about 2 months ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

context-degradation received a trust score of 86/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.

SkillShield's automated analysis identified 2 findings: 0 critical, 0 high, 2 medium, and 0 low severity. Key findings include Unsafe deserialization / dynamic eval, Incomplete Python script prevents full security 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 11, 2026 (commit 7942df36). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

93%

Static Code Analysis

93%

Dependency Graph

100%

LLM Behavioral Safety

100%

Behavioral Risk Signals

Filesystem Write

1 finding

Shell Execution

2 findings

Dynamic Code

2 findings

Security Findings2

Severity	Finding	Layer	Location
MEDIUM	Unsafe deserialization / dynamic eval Decryption followed by code execution Remove obfuscated code execution patterns. Legitimate code does not need base64-encoded payloads executed via eval, encrypted-then-executed blobs, or dynamic attribute resolution to call system functions.	Manifest	skills/context-degradation/scripts/degradation_detector.py:4
MEDIUM	Incomplete Python script prevents full security analysis The file `scripts/degradation_detector.py` is truncated. Specifically, the implementation of the `PoisoningDetector` class's `extract_claims` method and any subsequent code in the file are missing. This prevents a complete security assessment of the module, as potential vulnerabilities (e.g., command injection, data exfiltration, or credential harvesting) could exist in the unseen portions of the code. Without the full source, the security posture of this component cannot be fully guaranteed. Provide the complete source code for `scripts/degradation_detector.py` to enable a thorough and accurate security review.	Static	scripts/degradation_detector.py:170

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/bc42d32cea826a0d.svg)](https://skillshield.io/report/bc42d32cea826a0d)