The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the agent to execute shell commands for document compilation using the typst CLI, including the use of the --root .. flag to access files in parent directories as described in SKILL.md and references/common-patterns.md.
[COMMAND_EXECUTION]: It provides instructions for environment setup, such as creating Python virtual environments with uv venv and establishing symbolic links with ln -s in references/tool-routing.md to handle specific filesystem requirements for rendering charts.
[COMMAND_EXECUTION]: The skill includes shell scripts in references/data-driven.md that utilize bash and jq to automate the generation of multiple documents from structured CSV or JSON files.
[EXTERNAL_DOWNLOADS]: The skill makes use of numerous third-party packages from the Typst Universe repository (e.g., @preview/cetz, @preview/fletcher), which are retrieved by the Typst compiler during the document build process.
[PROMPT_INJECTION]: The skill defines patterns for ingesting data from external sources like JSON, CSV, and YAML files (references/data-driven.md) without explicit boundary markers or sanitization logic. This untrusted data is processed in a context that includes shell execution capabilities, presenting a surface for indirect prompt injection.

typst