The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided descriptions to generate executable HTML and JavaScript code, which represents a potential attack surface for indirect prompt injection.
Ingestion points: User descriptions of everyday objects are ingested via the agent's prompt in SKILL.md.
Boundary markers: The skill does not define explicit delimiters or instructions for the agent to ignore potential commands embedded within the user descriptions.
Capability inventory: The skill workflow utilizes a local Python script (scripts/build.py) and recommends an external web browser (google-chrome or chromium) to process and verify the generated output.
Sanitization: There are no explicit sanitization or validation routines provided to the agent for filtering user input before it is used to generate logic or geometry.
[COMMAND_EXECUTION]: The skill instructs the agent to execute local shell commands as part of its normal operation.
Evidence: SKILL.md contains instructions for the agent to run scripts/build.py using Python and to use a headless browser for rendering verification. These operations are intended and necessary for the skill's primary function of generating and validating interactive figures.

colorfigures