The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external YAML and Markdown files to define agent behavior, creating a surface where instructions within those files could influence the agent's actions.
Ingestion points: Reads user-provided .yaml workflow files and {role}.md agent definition files as described in SKILL.md.
Boundary markers: None specified for the content extracted from the external files.
Capability inventory: Utilizes Read and Write tools for file operations, the Bash tool for directory checks, and the Agent tool to spawn sub-agents.
Sanitization: No sanitization or schema validation of the content within the ingested files is mentioned.
[EXTERNAL_DOWNLOADS]: The skill recommends the installation of a supporting repository and package hosted by the author.
Evidence: Suggests users install agency-agents-zh via git clone https://github.com/jnMetaCode/agency-agents-zh.git or npm install agency-agents-zh if role directories are not found.
[COMMAND_EXECUTION]: Employs shell commands to perform environment checks during execution.
Evidence: Uses bash test -d to verify the existence of agent definition directories across several potential paths.

workflow-runner