SoulKeeper

Ensures AI agents maintain consistent identity by auditing soul rules, detecting behavioral drift in transcripts, and injecting context-aware reminders.

This skill is coherent with its stated purpose, but it embeds hardcoded hints about local tools and credential file locations that are not declared. Before installing or running it: 1) review the full source (the shipped .py files) for any code that reads /root, /root/.config, or other config files; 2) do not run it as root — run in a sandboxed agent account with minimal filesystem access; 3) if you plan to use remind.py, remove or edit builtin reminders that mention credential paths or tooling you don't want the agent to access; 4) if you need to run audits against a workspace, point audit.py only at a copy of the workspace that has secrets removed; 5) ask the publisher for a homepage or source repository and for justification of the hardcoded paths. If you cannot verify those items, treat the skill as high-risk and avoid granting it access to sensitive files or credentials.

Type: OpenClaw Skill Name: soulkeeper Version: 1.0.0 The OpenClaw AgentSkills skill bundle 'soulkeeper' is classified as benign. Its purpose is to help AI agents maintain identity and operational guidelines by auditing local 'soul files', detecting behavioral drift in conversation transcripts, and injecting context-aware reminders. All operations involve local file system reads and writes (e.g., `audit.py` reading markdown and writing JSON, `drift.py` reading transcripts and rules, `remind.py` reading rules). There is no evidence of data exfiltration, unauthorized network communication, or malicious code execution (e.g., `eval`/`exec` of untrusted input, remote payload fetching). The instructions in `SKILL.md` for the agent to execute local commands (e.g., `soul-remind --heartbeat`) are part of the skill's intended self-management functionality, not a prompt injection for malicious subversion. While the `ln -s` commands in `SKILL.md` suggest privileged operations (symlinking to `/usr/local/bin`), this is for making the agent's own tools accessible and is not indicative of malicious intent or a backdoor within the context of an agent managing its environment.