The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute the typst CLI for document compilation and runs a local verification script (scripts/verify-typst.sh) to check the environment and CJK font availability.
[DYNAMIC_EXECUTION]: The core workflow involves the agent dynamically generating Typst (.typ) source files based on user requirements and then executing these files via the typst compile command. This represents runtime compilation of agent-generated code.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from user prompts and external files (JSON, CSV) to generate documents, creating a vulnerability surface where untrusted data could influence the output.
Ingestion points: User instructions for document content (processed by SKILL.md) and external data sources loaded via functions like json() or csv() in generated .typ files (as seen in references/typst-templates.md).
Boundary markers: Absent; there are no specific delimiters or instructions in SKILL.md to ignore embedded instructions in user-provided data.
Capability inventory: Shell access via the Bash tool to run the typst compiler (referenced in SKILL.md) and file system access via the Write tool (referenced in SKILL.md).
Sanitization: Absent; the instructions and templates do not implement validation or escaping for interpolated user content.

typst