Personality Switcher

Create and switch between AI assistant personalities. Use /personality to list and activate saved personalities. Use /create-personality to design new personas with auto-filled SOUL and IDENTITY. Personalities persist across session boundaries and conversation compacting with automatic heartbeat restoration. Atomic switching with backup and rollback safeguards. Always backs up current state before switching.

This skill appears to do what it says (manage, persist, backup, and switch assistant personalities) and contains only local scripts — no remote downloads or hidden network calls were found. Before installing: 1) Inspect ~/.openclaw/openclaw.json (gateway config) and back it up — the installer will write to channels.telegram.customCommands to add /personality etc. and the unregister script will try to remove them on uninstall. 2) Expect a line calling restore_personality.py to be added to HEARTBEAT.md so the skill can reapply the active personality on each heartbeat; back up HEARTBEAT.md if you care about its contents. 3) Back up any important SOUL.md / IDENTITY.md files; the skill stores backups under ~/.openclaw/workspace/personalities/backups and leaves the personalities folder intact on uninstall. 4) If you run on a shared system or have sensitive tokens in the gateway config, review the register_telegram_commands.py/unregister_telegram_commands.py scripts to ensure the modifications are acceptable. 5) If you prefer manual control, run the register script yourself rather than allowing the installer to modify configs automatically. The behavior is coherent with the stated purpose; proceed if you accept the system-wide config and heartbeat changes.

Type: OpenClaw Skill Name: personality-switcher Version: 0.1.3 The skill bundle is classified as benign. While it performs high-privilege actions such as modifying the `HEARTBEAT.md` file for personality persistence (`install.sh`, `uninstall.sh`, `scripts/restore_personality.py`) and extensive file system operations (copying, renaming, deleting files and directories within `~/.openclaw/workspace/personalities/`), these actions are explicitly documented in `SKILL.md` and are central to the skill's stated purpose of managing AI personalities. All file operations are confined to the OpenClaw workspace and its designated subdirectories, and there is no evidence of data exfiltration, malicious execution, obfuscation, or prompt injection against the agent within the skill's own markdown instructions.