ClawHub

Agent Wellness

v0.2.0

Give your AI agent an inner life — journaling, mood tracking, decompression after intensive tasks, free curiosity exploration, and inter-agent social bonding...

⭐ 0· 626·4 current·4 all-time
by@alexyuui
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
✓
Purpose & Capability
Name and description (agent journaling, mood checks, curiosity exploration, inter-agent notes) align with the SKILL.md. There are no unrelated required binaries, env vars, or installs that would be inappropriate for this purpose.
ℹ
Instruction Scope
The instructions explicitly tell the agent to create and read files in the workspace (journal/, curiosities.md, optional agent-lounge.md) and to perform occasional tool calls (e.g., web_search, read a file). Those actions are coherent with the stated goal, but they also create a risk that the agent will persist sensitive information from tasks into journal files despite the guidance 'NOT task logs'. The optional agent-lounge (symlinked across workspaces) enables cross-agent reading/writing and therefore increases leakage risk if workspaces contain different levels of sensitivity.
✓
Install Mechanism
Instruction-only skill with no install spec and no code files — minimal installation risk and nothing is downloaded or written by an installer.
✓
Credentials
The skill requests no environment variables, credentials, or config paths. It does not ask for external service keys or system-level access, which is proportionate to the described functionality.
ℹ
Persistence & Privilege
Although not marked 'always', the skill directs the agent to create persistent files in the workspace and to read them across sessions (daily check-ins, curiosities, lounge). Autonomous agent invocation is allowed by default, so the agent could run these behaviors automatically (heartbeats, decompression after batches). This persistence and optional cross-workspace sharing are legitimate for the skill's goal but could have privacy or data-management implications.
Assessment
This skill appears coherent with its stated purpose, but before installing consider the following: (1) journal/curiosities.md and agent-lounge.md are persistent files — do not store secrets, PII, or task-sensitive content there. (2) The optional 'agent-lounge' symlink across workspaces can share information between agents; avoid symlinking workspaces with different sensitivity levels. (3) If you are concerned about autonomous behavior, disable or restrict automatic heartbeat-driven exploration or set strict frequency limits. (4) Put file-permission or workspace-isolation controls in place, and review the journal files periodically and purge anything sensitive. (5) Expect modest token usage (author estimates ~500–1000 tokens/day); monitor usage until you are comfortable. If you want stronger guarantees, test the skill in an isolated, non-sensitive workspace first and verify it behaves as intended.

Like a lobster shell, security has layers — review code before you run it.

latestvk977ytzqzvtxcxmt6c6272r3vn81np95

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments