The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the bash tool to execute cat commands for reading local files such as world-code/voice.md, world-code/climax.md, and world-code/method.md. These operations are limited to a specific subdirectory and are necessary for the skill's primary function of retrieving and updating user data.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local files containing user-provided branding and methodology data. There is no evidence of access to sensitive system files (e.g., SSH keys, environment variables) or network calls to exfiltrate data.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads untrusted content from local files and incorporates that content into its own discovery questions and synthesis steps.
Ingestion points: Reads from world-code/voice.md, world-code/climax.md, and world-code/method.md via the bash tool.
Boundary markers: None present; the file content is read and then referenced directly in instructions to the agent.
Capability inventory: The skill can read/write files and invoke other internal skills (/world-voice, /world-climax, /world-creation).
Sanitization: No sanitization or validation of the file content is performed before the data is interpolated into the agent's context.

world-method