The Agent Skills Directory

[NO_CODE]: The skill is entirely composed of Markdown files. No Python scripts, JavaScript files, or other executable binaries are present in the skill package.- [PROMPT_INJECTION]: The skill architecture is susceptible to indirect prompt injection.
Ingestion points: The skill is designed to process external, untrusted data including App Store reviews and information retrieved via the WebSearch and Read tools during competitor analysis.
Boundary markers: The skill files do not provide any delimiters or specific instructions for the agent to distinguish between data and potential commands within the processed text.
Capability inventory: The skill includes access to powerful tools such as Read, Write, Edit, and WebSearch. If an attacker-controlled review or web page contains malicious instructions, the agent could be manipulated into performing unintended file or network operations.
Sanitization: There are no instructions for the agent to validate, sanitize, or filter external content before it is incorporated into the generation process.

app-store