The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's operational workflow, described in agents/office-docs.md, involves generating temporary TypeScript scripts from templates and executing them using npx tsx. This is the intended primary purpose of the skill to produce document outputs.
[EXTERNAL_DOWNLOADS]: The scripts/verify-deps.sh script and the agent's setup instructions perform package management tasks via npm or pnpm. These operations target the official npm registry to manage well-known document generation libraries (docx, xlsx, pdf-lib, pptxgenjs).
[INDIRECT_PROMPT_INJECTION]: The skill contains a vulnerability surface where user-supplied content is interpolated into TypeScript templates before execution. While this facilitates document generation, it relies on the agent's logic to properly handle user input during the script-writing phase.
Ingestion points: User-provided data and text content for document generation (processed by agents/office-docs.md).
Boundary markers: None explicitly present in the template modification instructions.
Capability inventory: File writing and subprocess execution via npx tsx and npm install in the office-docs.md workflow.
Sanitization: Relies on the agent to adapt templates minimally and replace placeholders; no explicit sanitization logic is provided in the skill scripts themselves.

office