Detect credential theft and other security risks in AI skill definitions before installation.
npx skills add https://github.com/JNZader/javi-ai --skill skillguard-jnzader
Please help me install this Agent Skill. Skill: skillguard Source: https://github.com/JNZader/javi-ai/tree/main/own/skills/skillguard Command: npx skills add https://github.com/JNZader/javi-ai --skill skillguard-jnzader
Credential-aware pre-install reviews for AI skills to prevent credential leakage, code execution risks, data exfiltration, and scope escape before installation.
Run SkillGuard on the target skill to generate a safety report prior to installation.
High-intent search queries and answers about installing and using this skill.
To run a pre-install security scan, execute the scanner on the target AI skill definition to generate an auditable risk report with remediation guidance prior to installation.
Static analysis for AI skill vulnerability assessment is the process of detecting credential theft, code injection, data exfiltration, and scope escape by enforcing heuristic checks on skill definitions before installation.
Yes, you can automate pre-install security checks during skill installation workflows to intercept malicious patterns, network access attempts, and self-modification routines before the skill is added to your environment.
Pre-install skill vetting detects credential leakage risks by applying static and heuristic checks to identify credential patterns, network calls, and self-modification within AI skill definitions.
Yes, pre-install security scanning generates an auditable risk report with remediation guidance, providing a safety assessment for third-party skill vetting and credential-aware reviews.
Heuristic checks for AI skill scope escape detection are limited to static pattern analysis and known network access routines, meaning they enforce pre-install safety but may not catch novel obfuscation techniques.
pr-pmGuide prpm.json manifest structure for single and multi-package PRPM repositories.
acmacalisterParse and classify vulnerability reports with Bugcrowd VRT mapping and remediation steps.
cc01ccCreate, update, and validate Kilo SKILL.md files and kilo.jsonc configurations.
djacobsmeyerConfigure Claude Code sandbox policies in settings.json for secure code execution.
