Trust Assessment
cron-mastery received a trust score of 82/100, 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, 1 high, 1 medium, and 0 low severity. Key findings include Potential Data Exfiltration via Cron Job Listing, Agent Granted Broad Deletion Permissions for Cron Jobs.
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 12, 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 | |
|---|---|---|---|---|
| HIGH | Potential Data Exfiltration via Cron Job Listing The 'Janitor Prompt' instructs the agent to 'List all cron jobs including disabled ones' and 'Log what you deleted.' If cron job names or payloads contain sensitive user data (e.g., API keys, personal information, project details), this instruction could lead to the exfiltration of that data when the agent lists or logs the job details. While the janitor's intent is cleanup, the act of listing and logging all job details creates a vector for data exposure if sensitive information is stored within cron job definitions. Ensure that cron job names and payloads are sanitized or do not contain sensitive information. Implement stricter access controls or redaction mechanisms for `cron:list` and logging outputs. Consider if the janitor truly needs to log *all* details of deleted jobs, or just their IDs. | LLM | SKILL.md:60 | |
| MEDIUM | Agent Granted Broad Deletion Permissions for Cron Jobs The 'Janitor Prompt' instructs the agent to 'delete them' (cron jobs) based on specific criteria (`enabled: false` and `lastStatus: ok`). While the prompt includes a safeguard ('Do not delete active recurring jobs'), granting an LLM agent the ability to delete system-level jobs based on its interpretation of natural language criteria introduces a risk. A misinterpretation of the criteria, potentially influenced by adversarial input in job names or payloads, could lead to the unintended deletion of active or critical cron jobs, causing service disruption. Implement robust validation and confirmation steps before deletion. Consider using a more structured or programmatic approach for deletion criteria instead of relying solely on natural language interpretation by the LLM. Ensure the `cron:delete` tool has granular permissions, perhaps requiring explicit confirmation for critical jobs or only allowing deletion of jobs created by the agent itself. | LLM | SKILL.md:60 |
Scan History
Embed Code
[](https://skillshield.io/report/04e66bedf2847958)
Powered by SkillShield