The Agent Skills Directory

[PROMPT_INJECTION]: No patterns were detected that attempt to override system instructions, bypass safety filters, or extract system prompts. The instructions are focused entirely on the creative analysis of fictional settings.
[DATA_EXFILTRATION]: No network operations (e.g., curl, wget) or attempts to access sensitive file paths (e.g., .ssh, .aws, .env) were found. The skill operates solely on the text provided by the user within the agent's context.
[REMOTE_CODE_EXECUTION]: The skill does not download external scripts or packages. There are no patterns involving remote command execution or piped shell commands.
[NO_CODE]: The skill consists entirely of Markdown instructions and YAML metadata. It does not include any scripts (Python, JavaScript, shell) or executable binary content.
[COMMAND_EXECUTION]: No shell commands or subprocess calls are present. The skill utilizes the standard AskUserQuestion tool for interactive user confirmation, which is a safe and intended platform behavior.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted user-provided text (fictional world descriptions).
Ingestion points: User-supplied creative writing content processed during the 'World Audit' steps.
Boundary markers: The skill uses structured headers and steps, but does not implement explicit delimiters for user input.
Capability inventory: No risky capabilities (file-system write, network access, or shell execution) exist within the skill.
Sanitization: No specific sanitization or filtering of user input is performed, but the lack of dangerous tools makes the risk negligible.

s4h-writing-worldbuilding