360 Secure SkillHub
360 安全技能中心

Operator Hardening Pack

该 Skill 作为顶层操作入口,把启动硬化、角色编排与交接规范整合为一套工作流:先硬化 AGENTS/SOUL/USER/HEARTBEAT/记忆等核心文件,再对复杂任务启用角色分离执行,按 references/pack-runbook.md 的模板执行交接,最后交付结果并记录决策与后续项。它只给出流程、护栏与输出契约,不含可执行脚本。

cwheeler67
cwheeler67
热度 36.92
最近更新 2026-05-11 17:44:03
分类:自动化依赖:纯提示词应用场景:工作流 / 模板 / 任务管理

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add operator-hardening-pack
Metadata
name
operator-hardening-pack
description
A unified operating workflow for hardening agent setups and running complex tasks reliably by combining bootstrap hardening, role orchestration, and handoff discipline.

Operator Hardening Pack

Use this skill as the top-level entry point when you need a reliable, safety-first agent operating system.

Includes

  • agent-bootstrap-hardening: tighten AGENTS/SOUL/USER/HEARTBEAT/memory structure
  • team-agents-orchestrator: role-separated execution (operator/researcher/builder/editor)
  • team-role-launchkit: reusable role prompts + handoff contract templates

When to use

  • New workspace setup
  • Existing workspace feels inconsistent/noisy
  • Complex multi-step tasks with quality drift
  • Need repeatable role handoffs across long tasks

Workflow (recommended)

  1. Harden foundation
    • Run bootstrap hardening on core files.
  2. Enable role mode for complex work
    • Activate orchestrator with explicit role boundaries.
  3. Enforce handoff contract
    • Use template from references/pack-runbook.md.
  4. Deliver + log
    • Final output + decisions/follow-ups in memory.

Guardrails

  • External/public actions require explicit approval policy.
  • Keep instructions concise and enforceable.
  • Promote only durable, useful memory.
  • Prefer no-change over low-value churn.

Output contract

Return:

  • hardening/operation summary
  • files/artifacts changed
  • unresolved risks
  • next best step