# Openclaw Mem Self Improvement

> Inspect experimental self-improvement consolidation surfaces. Use for read-only validation, goal readback, staged skill proposals, curator review, and system status.

- Skill: `phenomenoner/openclaw-mem-self-improvement` (Agent Skill)
- Install (CLI): `npx skillmds@latest add phenomenoner/openclaw-mem-self-improvement`
- Raw SKILL.md: https://api.skillmd.com/api/skills/phenomenoner/openclaw-mem-self-improvement/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: phenomenoner (https://skillmd.com/u/phenomenoner)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/phenomenoner/openclaw-mem-self-improvement

---


# Self-improvement Lab

## Surfaces

```bash
openclaw-mem surface validate --inventory <inventory> --receipt <receipt> --json
openclaw-mem goal status --file <goal> --json
openclaw-mem goal pack --file <goal> --json
openclaw-mem skill-capture propose --text <proposal> --out <artifact> --json
openclaw-mem skill-curator review --skill-root <workspace>/skills --no-write --json
openclaw-mem mem-system status --json
```

Goal and system status are read-only. Curator review is report-only. Skill capture may write only an explicit staged proposal. These surfaces do not patch live skills, schedule cron, change gateway/plugin configuration, or enable auto-continuation. Higher-authority operations remain manual approval territory.

## Verify

```bash
openclaw-mem mem-system status --json
python -m pytest tests/test_self_improvement_surface.py tests/test_skill_capture.py -q
git diff --check
```