Roleplay

Create persistent characters, run practice scenarios, and track progress across roleplay sessions with activation control and structured feedback.

This skill is coherent with its description, but it stores roleplay data (including potentially sensitive therapy/medical scenarios and session logs) in ~/roleplay/. Before installing: 1) Decide whether you’re comfortable with local persistence—if not, disable saving or regularly clean the folder. 2) Protect ~/roleplay/ (file permissions, encryption, or use an account that isolates this data). 3) If you plan to use therapy/clinical scenarios, recognize the assistant is not a licensed clinician: add explicit runtime rules to refuse clinical diagnoses, provide crisis-resource links, and escalate to emergency guidance if the user reports imminent harm. 4) Confirm you accept the skill’s persona rules (it already prohibits impersonating living people) and that you want the agent to be allowed to read/write your home directory. If any of these are unacceptable, do not enable persistence or require the skill be modified to meet your privacy/safety needs.

Type: OpenClaw Skill Name: roleplay Version: 1.0.0 The skill bundle is designed for roleplaying and character management, requiring file system operations (read, write, move) to store character profiles, session logs, and other data within the user's `~/roleplay/` directory. This file access is directly aligned with the stated purpose and is not used to access sensitive system files or exfiltrate data. The `safeguards.md` file explicitly outlines rules to prevent the AI agent from engaging in harmful or legally risky content, such as impersonating living individuals or facilitating fraud. There is no evidence of malicious intent, prompt injection for harmful objectives, or other high-risk behaviors across any of the files.