The Agent Skills Directory

[SAFE]: The skill implements a static validator that reads local markdown files and directory structures to verify compliance with best practices. No dangerous patterns such as remote code execution, credential harvesting, or data exfiltration were detected.
[COMMAND_EXECUTION]: The skill uses a local bash script (references/validate.sh) to perform its validation logic. The script uses standard utilities like grep, sed, and find to analyze file contents. These operations are restricted to reading the skill directories and do not involve executing content from the files being analyzed.
[EXTERNAL_DOWNLOADS]: The documentation contains references to well-known organizations and services (Microsoft, OpenAI, Anthropic, Trail of Bits) for educational purposes. No automated downloads or script executions from these sources occur at runtime.

cali-skill-validator