The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains behavioral instructions directing the agent to render visualizations immediately and skip setup questions for efficiency. These instructions are focused on the specific formatting task and do not attempt to override safety filters or general agent guidelines.
[DATA_EXFILTRATION]: The skill configuration allows the use of file system discovery tools (Read, Grep, Glob) to gather context for diagrams. However, the skill does not include network access or any mechanisms for transferring data externally, and its output is strictly intended for inline chat responses.
[COMMAND_EXECUTION]: Documentation files mention various shell scripts and build tools used for development and testing (e.g., scripts/validate-tokens.sh, npm run test:ascii). These tools are not authorized for agent use in the skill's configuration and are not part of the runtime execution path.

glyph