The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands through Bash(cmd:*) to perform repository maintenance tasks. Specifically, it runs jq for JSON validation, executes pnpm run sync-marketplace to update catalog files, and triggers a local validation script ./scripts/validate-all-plugins.sh. These operations are localized to the repository environment and are consistent with the skill's purpose of automating plugin creation and validation.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests user-provided data (plugin name, description, keywords) to generate code and configuration files. 1. Ingestion points: User input for plugin metadata is gathered and interpolated into plugin.json, README.md, and markdown component files as described in SKILL.md. 2. Boundary markers: The instructions do not explicitly mandate delimiters or escaping for user-supplied strings during interpolation. 3. Capability inventory: The skill has Write access to create files and Bash access to execute scripts as noted in SKILL.md and references/ARD.md. 4. Sanitization: The skill includes mitigation steps such as enforcing kebab-case for names, validating JSON structure with jq, and running a comprehensive validation script before completion.

plugin-creator