本文へスキップ
TakoAPI
スキル一覧に戻る

error-recovery-automation

konscious0beast 作

Coding Agents & IDEs
504ダウンロード数はclawskills.shが公開している値です
0閲覧数
0件のいいねサインインするとこのスキルにいいねできます
—まだ評価がありません
あなたの評価

概要

Standardize handling of common OpenClaw errors (gateway restart, browser service unavailable, cron failures)

説明

このスキルでできること

Standardizes detection and recovery of common OpenClaw failures: gateway unresponsiveness, browser service outages, cron scheduler issues, and memory search errors. Provides bash script templates and cron scheduling patterns to automate restart sequences with verification steps. Includes escalation logic when automated recovery fails.

Reduces manual intervention by replacing ad-hoc restarts with structured scripts that detect, recover, verify, and escalate in a consistent sequence.

使いどころ

  • Automatically restart a crashed gateway without manual SSH
  • Schedule a browser health check every 30 minutes via cron
  • Rebuild the better-sqlite3 native module when memory search breaks
  • Escalate unrecoverable failures to a task inbox for manual review
  • Run a combined keep-alive script across all OpenClaw services

ワークフロー例

入力

User asks: set up automated browser recovery that runs every 30 minutes

エージェント
  1. 1

    Write browser-recovery.sh using the template: check status, stop, start, re-check

  2. 2

    Test the script manually by stopping the browser service and running the script

  3. 3

    Verify recovery with openclaw browser status --json

  4. 4

    Schedule the script as a cron job with openclaw cron add at 30-minute intervals

  5. 5

    Configure the job to announce only on failure via the delivery channel

出力

A scheduled cron job that automatically recovers the browser service and notifies on persistent failure

インストール

clawhub install konscious0beast/error-recovery-automation

CLIコマンドを実行するか、Claude Code、Cursor、Codexなどお使いのコーディングエージェントにプロンプトを貼り付けてください。

リンク