The Agent Skills Directory

[COMMAND_EXECUTION]: The skill operates by executing the unity-mcp-cli command via the shell to perform its tasks.
[DATA_EXFILTRATION]: The skill provides a way to read arbitrary files via the filePath parameter. This can be used to access sensitive files like credentials, environment variables, or private keys if the agent's input is not strictly validated or restricted to a specific directory.
[EXTERNAL_DOWNLOADS]: The documentation suggests installing unity-mcp-cli via the npm registry. This is an external dependency used to execute the skill's primary functions and is maintained by the skill's author.
[INDIRECT_PROMPT_INJECTION]: The skill ingests content from external files and returns it directly to the agent's context, creating a surface for malicious instructions to be processed.
Ingestion points: Content of the file at the path specified in the filePath parameter.
Boundary markers: Absent; the skill does not provide delimiters or warnings to the agent to treat the read content as untrusted data.
Capability inventory: The skill performs file system read operations using the unity-mcp-cli tool.
Sanitization: Absent; file data is returned as a raw string without filtering or escaping.

script-read