The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill automates the installation of system and Python dependencies required for rendering. These are standard packages from well-known official registries.\n
Evidence: The scripts/check_deps.sh script installs reportlab, weasyprint, markdown, and pygments via pip3, and pango, cairo, gdk-pixbuf, and libffi via brew on macOS.\n- [COMMAND_EXECUTION]: The skill relies on shell commands to execute its core conversion logic, dependency management, and configuration updates.\n
Evidence: SKILL.md triggers the Bash tool to run scripts/check_deps.sh and scripts/md_to_pdf.py, and it uses shell redirection to write JSON configuration files.\n- [PROMPT_INJECTION]: The skill's CONVERT+PROMPT mode, which applies LLM-based transformations to document content, creates an attack surface for indirect prompt injection.\n
Ingestion points: The skill reads untrusted external Markdown files using the Read tool.\n
Boundary markers: The instructions do not define delimiters to separate the content of the Markdown file from the agent's instructions during the pre-processing stage.\n
Capability inventory: The agent has access to Bash, Write, and Edit tools, which could be abused if the model is misled by instructions embedded within a processed file.\n
Sanitization: No sanitization or filtering is applied to the Markdown content before it is processed by the model.

md-to-pdf