The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local commands yq and php -l to provide functionality.
The SessionStart hook uses yq to read .coding_standards from .claude/config.yaml.
The PostToolUse hook uses php -l to check the syntax of PHP files after they are edited.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it reads and displays content from project files.
Ingestion points: .claude/config.yaml (read by yq) and any PHP file edited (processed by php -l).
Boundary markers: None present; the skill outputs configuration data directly to the agent's context.
Capability inventory: The skill has command execution capabilities via hooks defined in SKILL.md.
Sanitization: No explicit sanitization or instruction-ignoring delimiters are used for the echoed output from yq.

coding-standards