agent-hooksClaude Code Stop/PreToolUse/PostToolUse hook 脚本集。当 agent 提前停止、工具反复失败、或以投机语言完成任务时自动干预。每个脚本是独立的 bash hook，配到 settings.json 即可生效。不用于设计决策（用 harness-design-patter...
Install via ClawdBot CLI:
clawdbot install lanyasheng/agent-hooksGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Calls external URL not in known-safe list
https://code.claude.com/docs/en/hooksAudited Apr 16, 2026 · audit v1.0
Generated Apr 19, 2026
A software development team uses the agent to refactor a large legacy codebase. The agent may stop prematurely due to complexity, but Ralph hooks ensure it continues execution until the task is complete, while tool-error hooks prevent infinite retries on failed commands like linting or testing.
A data engineer employs the agent to debug a failing ETL pipeline. The agent might speculate on fixes without evidence, but doubt-gate hook forces it to provide concrete logs or error outputs, and post-edit-check hook validates any script edits for syntax errors immediately.
A DevOps team automates cloud infrastructure deployment using the agent. Tool-error hooks track repeated failures in commands like Terraform apply, blocking retries after five attempts to avoid resource conflicts, and Ralph hooks prevent early stops during long-running provisioning tasks.
A QA team uses the agent to run automated test suites across multiple environments. The agent could stop after partial test execution, but Ralph hooks enforce completion, while post-edit-check hook ensures any test script modifications are error-free before execution.
A marketing team leverages the agent to generate and edit technical documentation. The agent might produce speculative content, but doubt-gate hook requires evidence-based claims, and tool-error hooks manage failures in formatting or publishing tools to maintain workflow efficiency.
Integrate the hooks into a cloud-based AI development platform as a premium feature. Offer tiered subscriptions based on usage limits, targeting enterprises needing reliable agent execution for automated workflows, with revenue from monthly fees.
Provide consulting services to businesses for customizing and deploying these hooks in their specific AI agent environments. Charge per project or hourly for setup, training, and ongoing support, focusing on industries like fintech or healthcare.
Distribute the hooks as open-source software under MIT license to build community adoption. Generate revenue by offering enterprise support packages, including priority bug fixes, security audits, and dedicated assistance for large-scale deployments.
💬 Integration Tip
Ensure all hook scripts are placed in accessible paths and test them in a sandbox environment before production to avoid disrupting critical agent workflows.
Scored Apr 19, 2026
Meta-skill for AI agent self-improvement. Analyzes runtime logs to detect error patterns, regressions, and inefficiencies, then generates structured improvem...
Stop waiting for prompts. Keep working.
Turn OpenClaw into a learning-loop agent with seeded workspace rules, skill promotion, reflective memory, and proactive maintenance.
Local Python orchestration skill: multi-agent workflows via shared blackboard file, permission gating, token budget scripts, and persistent project context....
Meta-agent skill for orchestrating complex tasks through autonomous sub-agents. Decomposes macro tasks into subtasks, spawns specialized sub-agents with dynamically generated SKILL.md files, coordinates file-based communication, consolidates results, and dissolves agents upon completion. MANDATORY TRIGGERS: orchestrate, multi-agent, decompose task, spawn agents, sub-agents, parallel agents, agent coordination, task breakdown, meta-agent, agent factory, delegate tasks
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.