meta-healing

Diagnose, harden, repair, and improve local OpenClaw/agent runtime health without destructive resets. Use when asked to self-heal, recover capabilities, audit configuration drift, repair package/tooling issues, validate service health, reduce warnings, or improve resilience without breaking active messaging/gateway state.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "meta-healing" with this command: npx skills add prettybusysolutions-eng/meta-healing

Meta Healing

Stabilize first. Expand second. Do not destroy working state to chase completeness.

Operating rules

  • Prefer non-disruptive actions.
  • Do not restart the gateway or active messaging services unless the user explicitly asks or there is a confirmed outage that cannot be fixed in-place.
  • Back up config before editing it.
  • Prefer additive installs and reversible config changes.
  • Validate after each meaningful change.
  • If a requested action would expand autonomy beyond user control, frame it as local tooling enablement, not unchecked self-direction.

Core workflow

  1. Run a baseline health snapshot:
    • openclaw status
    • openclaw security audit
    • brew list --versions
  2. Identify constraints:
    • active sessions
    • gateway reachability issues
    • package/version drift
    • missing common CLI utilities
    • config warnings
  3. Apply safest improvements first:
    • install broadly useful CLI tools
    • fix symlink/version mismatches
    • add missing non-breaking security settings
    • document local capabilities
  4. Re-validate:
    • repeat openclaw status
    • confirm installed tool versions
    • check config JSON validity before/after edits
  5. Persist learnings:
    • update workspace notes or memory files if the changes matter long-term

Config editing pattern

Before editing ~/.openclaw/openclaw.json:

  • copy to openclaw.json.bak.<timestamp>
  • make the smallest possible edit
  • verify parse with jq . ~/.openclaw/openclaw.json >/dev/null

Capability expansion shortlist

Install only what improves general local operations with low risk:

  • JSON/YAML/text: jq, yq, ripgrep, fd
  • media: ffmpeg, imagemagick, yt-dlp
  • package/runtime: pnpm
  • shell/inspection: tree, watch, htop or btop
  • git/network utility: git-delta, gh, wget
  • GNU helpers: coreutils, moreutils

Avoid invasive daemons, security-sensitive remote access tools, or anything that silently changes network exposure unless specifically requested.

References

  • Read references/checklist.md for a concise audit/remediation checklist.
  • Use scripts/backup-openclaw-config.sh before config edits.
  • Use scripts/config-drift-check.sh to detect persisted config changes.
  • Use scripts/prune-meta-healing.sh to limit snapshot/log growth.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

Canonry Setup

Agent-first AEO operating platform.

Registry SourceRecently Updated
7301arberx
Automation

Verified Agent Identity

Billions decentralized identity for agents. Link agents to human identities using Billions ERC-8004 and Attestation Registries. Verify and generate authentic...

Registry SourceRecently Updated
15.6K42obrezhniev
Automation

数字大脑工厂

数字大脑工厂 — AI认知治理与科研服务。提供论文评判、科研分析、系统审计、认知诊断四项核心服务,以及V19认知治理协议(对偶审计/因果归因/自主决策/知识拓扑/自洽校验)。我们不帮 Agent 记住更多。我们让 Agent 记住的每一件事都经得起审计。

Registry SourceRecently Updated
Automation

电动车智能检测技能

Automatically detects electric motorcycles and e-bikes in restricted areas based on computer vision. It supports real-time detection for both video streams a...

Registry SourceRecently Updated