The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the rg (ripgrep) utility to perform sanity checks by searching for template names within the project files.
[DATA_EXFILTRATION]: While no external network exfiltration is present, the skill reads sensitive template configuration files from the src/templates/ directory to update other project files.
[INDIRECT_PROMPT_INJECTION]: The skill reads template names from source code files (src/templates/*.ts) and interpolates them directly into other agent instructions (SKILL.md) and prompt files (Prompt.ts, prompt.md).
Ingestion points: Template names are collected from object keys in src/templates/*.ts and src/templates/built-in.ts (SKILL.md).
Boundary markers: Absent. There are no instructions to sanitize or validate the template names before they are used to update prompt files.
Capability inventory: The skill has the capability to write to multiple files including .skills/infographic-creator/SKILL.md, site/src/components/AIPlayground/Prompt.ts, and .skills/infographic-syntax-creator/references/prompt.md. It also executes the rg command.
Sanitization: Absent. The instructions do not specify any validation for the template names, which could lead to command injection if a template name contains shell metacharacters used in the rg sanity check.

infographic-template-updater