Scan agent skill code and metadata for credential leakage, malware, and prompt injection.
npx skills add https://github.com/oakencore/skillvet --skill skillvet
Please help me install this Agent Skill. Skill: skillvet Source: https://github.com/oakencore/skillvet/tree/main Command: npx skills add https://github.com/oakencore/skillvet --skill skillvet
Skillvet provides automated security auditing for agent skills to prevent malware, credential theft, and unsafe behavior before installation.
Run the scanner on the target skill directory to generate a findings report.
High-intent search queries and answers about installing and using this skill.
To audit an agent skill for malware and credential leakage, run a security scanner on the target skill directory to check code, metadata, and setup, generating a findings report that flags unsafe patterns before installation.
Yes, you can detect prompt injection and obfuscation patterns automatically by scanning skills against an embedded rule set and pattern database designed to identify campaign-specific attack signatures and exfiltration attempts.
Skill vetting supports CI-friendly outputs by generating JSON and SARIF reports, allowing automated security audits to integrate directly into install, audit, or vet workflows without requiring full skill installation.
The best way to scan a public marketplace skill without installing it is to use remote scanning capabilities that analyze the skill's code and metadata to block unsafe behavior and prevent credential theft.
Security audits identify exfiltration attempts by applying comprehensive rule sets to scan target skill directories, checking code and setup configurations for known attack signatures and unauthorized data transmission patterns.
Running a security audit on a ClawHub skill requires access to the target skill directory and a scanner equipped with pattern databases to detect malware, credential leakage, and prompt injection without additional dependencies.
OliverOuyangEnforce governance for adding, updating, and deleting skill units.
leprachuanScan AI agent skills for security vulnerabilities and output severity-rated findings.
SjdjdiejdrirhdkjejRun post-implementation quality gates and generate reports before release.
voidrotConfigure SAST tools like Semgrep, CodeQL, and SonarQube for CI/CD pipelines.
