The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local command-line utilities such as ripgrep (rg), git, and perl to inventory and audit typographic dashes within the working tree and commit history.- [SAFE]: The skill implements a robust security model by instructing the agent to treat all scanned file contents, commit messages, and command outputs as data rather than instructions, which effectively mitigates the risk of indirect prompt injection.- [SAFE]: Enforcement mechanisms, including a git commit-msg hook and a platform-specific command matcher, are designed to run locally and provide deterministic validation of hyphen usage without external dependencies or network access.- [SAFE]: The use of Unicode hex escape sequences in regular expressions is technically required for the skill's primary function of detecting specific dash code points and is not an attempt at obfuscation.

dashfix