The Agent Skills Directory

[SAFE]: The skill is designed for strategic market analysis and does not contain any malicious logic or dangerous tool invocations. Analysis of the scripts and metadata confirms that operations are limited to text processing and generation.
[PROMPT_INJECTION]: The skill uses persona-setting instructions such as "You are now acting as the Masterpiece Competitor Analysis Engine" to define the agent's role. These instructions do not attempt to bypass platform safety filters or extract system prompts.
[DATA_EXPOSURE]: The skill reads project-specific configuration files like brand-voice.md and meta.json. There are no attempts to access sensitive system files (e.g., SSH keys, AWS credentials) or environment secrets.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input via the $ARGUMENTS parameter.
Ingestion points: User-provided competitor data and industry information are injected into the prompt via $ARGUMENTS in SKILL.md.
Boundary markers: No delimiters or instructions to ignore embedded commands are present around the user input.
Capability inventory: The skill does not have access to shell commands, network operations, or file-writing tools across its files.
Sanitization: No input validation or sanitization of user-supplied data is performed. Given the lack of dangerous capabilities, this risk is minimal.

competitor-analysis