The Agent Skills Directory

[SAFE]: The skill is entirely instructional and composed of Markdown files. It does not include any executable scripts (.sh, .py, .js), binaries, or configuration files that would perform actions on the host system.
[SAFE]: All referenced tools (e.g., eslint, ruff, prettier, pytest, gofmt) are standard, well-known development utilities. There are no references to unverifiable or suspicious third-party packages or remote scripts.
[SAFE]: The skill explicitly recommends security-positive practices in the 'Validator design checklist,' such as ensuring validators are deterministic and avoid network calls.
[SAFE]: No patterns of prompt injection, obfuscation, data exfiltration, or privilege escalation were detected. The skill's content matches its stated purpose of improving the reliability of AI-generated code changes.

agentic-validators