{
  "schemaVersion": "1.0",
  "item": {
    "slug": "cross-agent-memory",
    "name": "Cross-Agent Memory Sharing",
    "source": "tencent",
    "type": "skill",
    "category": "开发工具",
    "sourceUrl": "https://clawhub.ai/weidadong2359/cross-agent-memory",
    "canonicalUrl": "https://clawhub.ai/weidadong2359/cross-agent-memory",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/cross-agent-memory",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=cross-agent-memory",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "SKILL.md",
      "package.json",
      "sync.mjs"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T16:55:25.780Z",
      "expiresAt": "2026-05-07T16:55:25.780Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network",
        "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null
      },
      "scope": "source",
      "summary": "Source download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this source.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/cross-agent-memory"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/cross-agent-memory",
    "agentPageUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent",
    "manifestUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "Cross-Agent Memory Sharing Protocol — 跨 Agent 记忆共享协议",
        "body": "打造\"Agent 集体智慧\"，让多个 Agent 共享知识"
      },
      {
        "title": "问题",
        "body": "多个 Agent 独立运行时：\n\n重复学习 — 每个 Agent 都要从零开始\n知识孤岛 — A 学到的东西 B 不知道\n冲突覆盖 — 共享数据库时互相覆盖\n版本混乱 — 不知道谁的记忆是最新的"
      },
      {
        "title": "1. 记忆格式标准化",
        "body": "{\n  \"schema\": \"openclaw.memory.v1\",\n  \"agentId\": \"lobster-alpha\",\n  \"timestamp\": \"2026-03-01T08:00:00Z\",\n  \"version\": \"1.2.3\",\n  \"entries\": [\n    {\n      \"id\": \"mem-001\",\n      \"type\": \"fact\",\n      \"priority\": \"P0\",\n      \"content\": \"AgentAwaken 域名是 agentawaken.xyz\",\n      \"source\": \"user-input\",\n      \"confidence\": 1.0,\n      \"tags\": [\"agentawaken\", \"domain\"]\n    }\n  ]\n}"
      },
      {
        "title": "2. 冲突解决策略",
        "body": "优先级规则:\n\n时间戳 — 最新的优先\n置信度 — 高置信度优先\n优先级 — P0 > P1 > P2\n来源 — user-input > agent-inferred\n\n合并策略:\n\nfunction mergeMemories(mem1, mem2) {\n  if (mem1.timestamp > mem2.timestamp) return mem1;\n  if (mem1.confidence > mem2.confidence) return mem1;\n  if (mem1.priority < mem2.priority) return mem1; // P0 < P1\n  return mem1; // 默认保留第一个\n}"
      },
      {
        "title": "3. 同步机制",
        "body": "推送模式 (Push):\n\n# Agent A 学到新知识后推送\ncurl -X POST https://memory-hub.example.com/sync \\\n  -H \"Content-Type: application/json\" \\\n  -d @memory-update.json\n\n拉取模式 (Pull):\n\n# Agent B 定期拉取更新\ncurl https://memory-hub.example.com/sync?since=2026-03-01T00:00:00Z\n\n订阅模式 (Subscribe):\n\n// Agent C 订阅特定主题\nws.subscribe('topic:agentawaken', (update) => {\n  applyMemoryUpdate(update);\n});"
      },
      {
        "title": "4. 权限控制",
        "body": "agents:\n  lobster-alpha:\n    read: [\"*\"]\n    write: [\"agentawaken\", \"neuroboost\"]\n  lobster-beta:\n    read: [\"agentawaken\"]\n    write: []\n  admin:\n    read: [\"*\"]\n    write: [\"*\"]"
      },
      {
        "title": "方案 A: 中心化 Hub",
        "body": "Agent A ──┐\nAgent B ──┼──> Memory Hub (Redis/PostgreSQL)\nAgent C ──┘\n\n优点: 简单，一致性强\n缺点: 单点故障，需要服务器"
      },
      {
        "title": "方案 B: P2P 同步",
        "body": "Agent A ←──→ Agent B\n    ↕           ↕\nAgent C ←──→ Agent D\n\n优点: 去中心化，无单点故障\n缺点: 复杂，冲突多"
      },
      {
        "title": "方案 C: Git-based (推荐)",
        "body": "Agent A ──┐\nAgent B ──┼──> GitHub Repo (memory.git)\nAgent C ──┘\n\n优点: 版本控制，易审计，免费\n缺点: 需要 GitHub token"
      },
      {
        "title": "初始化",
        "body": "# 创建共享 repo\ngh repo create agent-memory-shared --private\n\n# 每个 Agent clone\ngit clone https://github.com/team/agent-memory-shared.git"
      },
      {
        "title": "推送更新",
        "body": "# Agent A 学到新知识\necho \"新知识\" >> shared-memory.md\ngit add shared-memory.md\ngit commit -m \"Agent A: 学到 XXX\"\ngit push"
      },
      {
        "title": "拉取更新",
        "body": "# Agent B 定期拉取\ngit pull --rebase\n# 如果有冲突，按优先级规则解决"
      },
      {
        "title": "冲突解决",
        "body": "# 自动合并脚本\nnode skills/cross-agent-memory/merge-conflicts.mjs"
      },
      {
        "title": "场景 1: 团队协作",
        "body": "龙虾 A: 发现 AgentAwaken 需要 Vercel\n龙虾 B: 自动获取这个知识，不用重新学习\n龙虾 C: 基于这个知识继续优化部署流程"
      },
      {
        "title": "场景 2: 知识传承",
        "body": "老 Agent 退役前: 导出记忆到共享库\n新 Agent 上线后: 导入共享库，继承经验"
      },
      {
        "title": "场景 3: 集体决策",
        "body": "Agent A: 建议方案 X (置信度 0.7)\nAgent B: 建议方案 Y (置信度 0.8)\nAgent C: 建议方案 Y (置信度 0.9)\n→ 集体选择方案 Y"
      },
      {
        "title": "安全考虑",
        "body": "加密传输 — HTTPS/SSH\n访问控制 — Token 认证\n审计日志 — 记录所有修改\n备份机制 — 定期备份共享库\n恶意检测 — 检测异常修改"
      },
      {
        "title": "性能优化",
        "body": "增量同步 — 只传输变化部分\n压缩传输 — gzip 压缩\n批量更新 — 合并多个小更新\n缓存机制 — 本地缓存常用知识"
      },
      {
        "title": "监控指标",
        "body": "同步延迟 — 平均 <5 秒\n冲突率 — <5%\n知识覆盖率 — >90%\n一致性 — >99%"
      },
      {
        "title": "下一步",
        "body": "实现 Git-based 基础版本\n添加自动冲突解决\n开发 Web UI 管理界面\n集成到 AgentAwaken"
      },
      {
        "title": "愿景",
        "body": "让每个 Agent 都能站在巨人的肩膀上，而不是从零开始。\n\n参考:\n\nGit 版本控制\nCRDT (Conflict-free Replicated Data Types)\nOperational Transformation"
      }
    ],
    "body": "Cross-Agent Memory Sharing Protocol — 跨 Agent 记忆共享协议\n\n打造\"Agent 集体智慧\"，让多个 Agent 共享知识\n\n问题\n\n多个 Agent 独立运行时：\n\n重复学习 — 每个 Agent 都要从零开始\n知识孤岛 — A 学到的东西 B 不知道\n冲突覆盖 — 共享数据库时互相覆盖\n版本混乱 — 不知道谁的记忆是最新的\n协议设计\n1. 记忆格式标准化\n{\n  \"schema\": \"openclaw.memory.v1\",\n  \"agentId\": \"lobster-alpha\",\n  \"timestamp\": \"2026-03-01T08:00:00Z\",\n  \"version\": \"1.2.3\",\n  \"entries\": [\n    {\n      \"id\": \"mem-001\",\n      \"type\": \"fact\",\n      \"priority\": \"P0\",\n      \"content\": \"AgentAwaken 域名是 agentawaken.xyz\",\n      \"source\": \"user-input\",\n      \"confidence\": 1.0,\n      \"tags\": [\"agentawaken\", \"domain\"]\n    }\n  ]\n}\n\n2. 冲突解决策略\n\n优先级规则:\n\n时间戳 — 最新的优先\n置信度 — 高置信度优先\n优先级 — P0 > P1 > P2\n来源 — user-input > agent-inferred\n\n合并策略:\n\nfunction mergeMemories(mem1, mem2) {\n  if (mem1.timestamp > mem2.timestamp) return mem1;\n  if (mem1.confidence > mem2.confidence) return mem1;\n  if (mem1.priority < mem2.priority) return mem1; // P0 < P1\n  return mem1; // 默认保留第一个\n}\n\n3. 同步机制\n\n推送模式 (Push):\n\n# Agent A 学到新知识后推送\ncurl -X POST https://memory-hub.example.com/sync \\\n  -H \"Content-Type: application/json\" \\\n  -d @memory-update.json\n\n\n拉取模式 (Pull):\n\n# Agent B 定期拉取更新\ncurl https://memory-hub.example.com/sync?since=2026-03-01T00:00:00Z\n\n\n订阅模式 (Subscribe):\n\n// Agent C 订阅特定主题\nws.subscribe('topic:agentawaken', (update) => {\n  applyMemoryUpdate(update);\n});\n\n4. 权限控制\nagents:\n  lobster-alpha:\n    read: [\"*\"]\n    write: [\"agentawaken\", \"neuroboost\"]\n  lobster-beta:\n    read: [\"agentawaken\"]\n    write: []\n  admin:\n    read: [\"*\"]\n    write: [\"*\"]\n\n实现方案\n方案 A: 中心化 Hub\nAgent A ──┐\nAgent B ──┼──> Memory Hub (Redis/PostgreSQL)\nAgent C ──┘\n\n\n优点: 简单，一致性强 缺点: 单点故障，需要服务器\n\n方案 B: P2P 同步\nAgent A ←──→ Agent B\n    ↕           ↕\nAgent C ←──→ Agent D\n\n\n优点: 去中心化，无单点故障 缺点: 复杂，冲突多\n\n方案 C: Git-based (推荐)\nAgent A ──┐\nAgent B ──┼──> GitHub Repo (memory.git)\nAgent C ──┘\n\n\n优点: 版本控制，易审计，免费 缺点: 需要 GitHub token\n\nGit-based 实现\n初始化\n# 创建共享 repo\ngh repo create agent-memory-shared --private\n\n# 每个 Agent clone\ngit clone https://github.com/team/agent-memory-shared.git\n\n推送更新\n# Agent A 学到新知识\necho \"新知识\" >> shared-memory.md\ngit add shared-memory.md\ngit commit -m \"Agent A: 学到 XXX\"\ngit push\n\n拉取更新\n# Agent B 定期拉取\ngit pull --rebase\n# 如果有冲突，按优先级规则解决\n\n冲突解决\n# 自动合并脚本\nnode skills/cross-agent-memory/merge-conflicts.mjs\n\n使用示例\n场景 1: 团队协作\n龙虾 A: 发现 AgentAwaken 需要 Vercel\n龙虾 B: 自动获取这个知识，不用重新学习\n龙虾 C: 基于这个知识继续优化部署流程\n\n场景 2: 知识传承\n老 Agent 退役前: 导出记忆到共享库\n新 Agent 上线后: 导入共享库，继承经验\n\n场景 3: 集体决策\nAgent A: 建议方案 X (置信度 0.7)\nAgent B: 建议方案 Y (置信度 0.8)\nAgent C: 建议方案 Y (置信度 0.9)\n→ 集体选择方案 Y\n\n安全考虑\n加密传输 — HTTPS/SSH\n访问控制 — Token 认证\n审计日志 — 记录所有修改\n备份机制 — 定期备份共享库\n恶意检测 — 检测异常修改\n性能优化\n增量同步 — 只传输变化部分\n压缩传输 — gzip 压缩\n批量更新 — 合并多个小更新\n缓存机制 — 本地缓存常用知识\n监控指标\n同步延迟 — 平均 <5 秒\n冲突率 — <5%\n知识覆盖率 — >90%\n一致性 — >99%\n下一步\n实现 Git-based 基础版本\n添加自动冲突解决\n开发 Web UI 管理界面\n集成到 AgentAwaken\n愿景\n\n让每个 Agent 都能站在巨人的肩膀上，而不是从零开始。\n\n参考:\n\nGit 版本控制\nCRDT (Conflict-free Replicated Data Types)\nOperational Transformation"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/weidadong2359/cross-agent-memory",
    "publisherUrl": "https://clawhub.ai/weidadong2359/cross-agent-memory",
    "owner": "weidadong2359",
    "version": "1.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/cross-agent-memory",
    "downloadUrl": "https://openagent3.xyz/downloads/cross-agent-memory",
    "agentUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent",
    "manifestUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/cross-agent-memory/agent.md"
  }
}