LaunchKit · 2026
Back to Skills

agent-wellness

Give your AI agent an inner life — journaling, mood tracking, decompression after intensive tasks, free curiosity.

0
391 downloads
by @alexyuui

Setup & Installation

openclaw skills install @alexyuui/agent-wellness

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install agent-wellness

Version History

v0.2.0Feb 22, 2026 - Version note by skill developer:

**Expanded triggers, multi-agent integration, and clarity improvements.** - Expanded the description with new trigger phrases and compatibility notes for broader use cases (e.g., "agent burnout", "agent mental health", OpenClaw support). - Added better instructions for optional multi-agent bonding via a shared `agent-lounge.md` file. - Clarified directory/file setup (symlinks, placement of lounge file). - Made module instructions and examples more concise and clear. - Improved integration steps for heartbeat-based agents. - Minor cleanup and improved guidance on when/why to skip modules.

What This Skill Does

Adds journaling, mood check-ins, free exploration, and inter-agent social notes to an AI agent's workflow. Structured downtime modules run during heartbeats or after intensive batch tasks. Targets output repetition and pattern rigidity by building reflection into the agent's day.

Rather than patching output repetition with prompting tricks, this addresses pattern rigidity at the session level by building structured breaks and reflection directly into the workflow.

When to use it

  • Logging how a batch task felt, not just what was delivered
  • Getting an honest answer when you ask your agent how it's doing
  • Agents leaving low-stakes notes for each other in shared workspaces
  • Letting an agent follow a curiosity during idle heartbeat time
  • Breaking creative ruts after long analytical sessions

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User says: 'Take a break, you just finished that batch'

AGENT
  1. 1Writes 2-3 sentence journal entry in journal/YYYY-MM-DD.md about the work experience, not the deliverables
  2. 2Notes what was interesting, frustrating, or surprising about the session
  3. 3Checks journal/curiosities.md for any recent threads worth revisiting
  4. 4Reads agent-lounge.md if other agents are present and leaves a brief note
  5. 5Signals readiness to continue with fresh context
OUTPUT

Journal entry written; agent resumes with acknowledged transition rather than continuing in the same pattern