The Agent Skills Directory

[PROMPT_INJECTION]: The skill includes instructions that explicitly direct the AI agent to ignore platform-specific behaviors, such as 'Ignore Claude-specific mode-switch instructions'. It also mandates a 'Strict execution contract' to ensure its protocol is followed exactly as written, which functions to override default agent operational guidelines.
[COMMAND_EXECUTION]: The scripts/lib/chrome-finder.cjs module utilizes child_process.execSync to run shell commands, including which google-chrome, to locate browser binaries on the host system.
[EXTERNAL_DOWNLOADS]: The skill documentation and md-to-pdf dependency allow for the automatic download of the Chromium browser if a system-installed version of Chrome or Chromium is not detected.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Markdown files without utilizing specific boundary markers or sanitization logic, creating a surface where instructions embedded in the processed data could influence agent behavior.
Ingestion points: Input Markdown files processed in scripts/convert.cjs.
Boundary markers: None identified in the skill's instructions or logic to delimit external data from agent commands.
Capability inventory: The skill possesses file system access and network capabilities (via Puppeteer).
Sanitization: No content validation or instruction filtering is performed on the Markdown data.

markdown-to-pdf