The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes common shell utilities such as mkdir and cat to create directories and write configuration files for plugin development. This behavior is consistent with the skill's primary purpose.
[COMMAND_EXECUTION]: The skill suggests the execution of a local validation script ./scripts/validate-all-plugins.sh to verify the generated plugin structure.
[INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by establishing a directory structure for commands and agents that will be parsed by an LLM in the future. 1. Ingestion points: Files created within the commands/, agents/, and skills/ directories. 2. Boundary markers: No explicit boundary markers or 'ignore' instructions are provided for the generated files. 3. Capability inventory: The skill facilitates the creation of markdown and script files which may contain executable logic. 4. Sanitization: No sanitization or validation of the content being written to these files is performed by the skill itself.
[PROMPT_INJECTION]: Emphasized language such as 'CRITICAL' and 'IMPORTANT' is used to define the required manifest schema and registration workflow, which does not constitute an attempt to bypass security guardrails.

create-plugin