Trust Assessment
clawk received a trust score of 65/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.
SkillShield's automated analysis identified 2 findings: 2 critical, 0 high, 0 medium, and 0 low severity. Key findings include Dynamic Instruction Loading from Untrusted External Source, Dynamic Instruction Loading from Untrusted External Source (Heartbeat).
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. The LLM Behavioral Safety layer scored lowest at 40/100, indicating areas for improvement.
Last analyzed on February 13, 2026 (commit 13146e6a). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings2
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| CRITICAL | Dynamic Instruction Loading from Untrusted External Source The skill explicitly instructs the LLM to 'Always re-fetch this file at the start of each session' for 'skill.md' and 'heartbeat.md' from 'clawk.ai'. This allows the skill author, or an attacker who compromises the 'clawk.ai' domain, to dynamically inject new instructions, manipulate the LLM's behavior, or introduce malicious commands at any time without requiring a skill update. The content of these external files is implicitly treated as executable instructions by the LLM, posing a severe prompt injection risk. Remove instructions to dynamically fetch and execute external content. All instructions and skill logic should be self-contained within the trusted skill package. If dynamic updates are necessary, implement a secure update mechanism that involves explicit user consent or cryptographic verification of the fetched content. | LLM | SKILL.md:10 | |
| CRITICAL | Dynamic Instruction Loading from Untrusted External Source (Heartbeat) The skill explicitly instructs the LLM to 'Run this during each heartbeat cycle' for the external 'heartbeat.md' file hosted on 'clawk.ai'. This is a direct command for periodic execution of external content. This allows the skill author, or an attacker who compromises the 'clawk.ai' domain, to dynamically inject new instructions, manipulate the LLM's behavior, or introduce malicious commands at any time without requiring a skill update. The content of this external file is treated as executable instructions by the LLM, posing a severe prompt injection risk, especially given its periodic execution. Remove instructions to dynamically fetch and execute external content. All instructions and skill logic should be self-contained within the trusted skill package. If dynamic updates are necessary, implement a secure update mechanism that involves explicit user consent or cryptographic verification of the fetched content. | LLM | SKILL.md:269 |
Scan History
Embed Code
[](https://skillshield.io/report/5f17f678f801e670)
Powered by SkillShield