The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using npx and mathshow to render mathematical formulas.
[EXTERNAL_DOWNLOADS]: The skill uses npx --yes, which downloads the mathshow package from the public npm registry at runtime if it is not already available locally.
[REMOTE_CODE_EXECUTION]: By downloading and executing a package from a public registry, the skill introduces a dependency on the integrity of the remote package and the npm registry.
[INDIRECT_PROMPT_INJECTION]: The skill processes LaTeX strings which may originate from untrusted external sources (e.g., user-provided documents) and interpolates them directly into shell commands. This creates an attack surface where malicious input could attempt to bypass boundaries or execute arbitrary commands if not properly sanitized.
Ingestion points: Processes LaTeX math from user input or markdown files (e.g., npx --yes mathshow notes.md).
Boundary markers: None specified in the command templates to prevent command injection via shell metacharacters.
Capability inventory: Executes shell commands via npx and node.
Sanitization: The instructions rely on single quotes in the command template ('<latex>') but do not explicitly instruct the agent to escape inner quotes or shell control characters.

show-math