当前位置: 首页 > article >正文

OpenClaw 解决运行一些漏洞

1.发现问题运行 openclaw status 发现3个严重的安全威胁$ openclaw status OpenClaw2026.3.13(61d171a)— The only crabinyour contacts you actually want to hear from. 11:54:50[plugins]feishu_doc: Registered feishu_doc, feishu_app_scopes11:54:50[plugins]feishu_chat: Registered feishu_chat tool11:54:50[plugins]feishu_wiki: Registered feishu_wiki tool11:54:50[plugins]feishu_drive: Registered feishu_drive tool11:54:50[plugins]feishu_bitable: Registered bitable tools │11:54:50[plugins]feishu_doc: Registered feishu_doc, feishu_app_scopes11:54:50[plugins]feishu_chat: Registered feishu_chat tool11:54:50[plugins]feishu_wiki: Registered feishu_wiki tool11:54:50[plugins]feishu_drive: Registered feishu_drive tool11:54:50[plugins]feishu_bitable: Registered bitable tools ◇ │ ◇ OpenClaw status Overview ┌─────────────────┬───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐ │ Item │ Value │ ├─────────────────┼───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┤ │ Dashboard │ http://127.0.0.1:18789/ │ │ OS │ linux6.8.0-71-generic(x64)·node24.14.0 │ │ Tailscale │ off │ │ Channel │ stable(default)│ │ Update │pnpm·npmlatest2026.3.13 │ │ Gateway │local· ws://127.0.0.1:18789(local loopback)· unreachable(missing scope: operator.read)│ │ Gatewayservice│ systemd installed · enabled · running(pid793469, state active)│ │ Nodeservice│ systemd not installed │ │ Agents │1·1bootstrapfilepresent · sessions2· default main active 14h ago │ │ Memory │0files ·0chunks · dirty · sources memory · plugin memory-core · vector unknown · fts ready · cache on(0)│ │ Probes │ skipped(use --deep)│ │ Events │ none │ │ Heartbeat │ 30m(main)│ │ Sessions │2active · default MiniMax-M2.5(200k ctx)· ~/.openclaw/agents/main/sessions/sessions.json │ └─────────────────┴───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘ Security audit Summary:3critical ·4warn ·1info CRITICAL Open groupPolicy with elevated tools enabled FoundgroupPolicyopenat: - channels.feishu.groupPolicy With tools.elevated enabled, a prompt injectioninthose rooms can become a high-impact incident. Fix: SetgroupPolicyallowlistand keep elevated allowlists extremely tight. CRITICAL Open groupPolicy with runtime/filesystem tools exposed FoundgroupPolicyopenat: - channels.feishu.groupPolicy Risky tool exposure contexts: - agents.defaults(sandboxoff;runtime[exec, process];fs[read, wri… Fix: Foropengroups, prefertools.profilemessaging(or deny group:runtime/group:fs),settools.fs.workspaceOnlytrue, and useagents.defaults.sandbox.modeallforexposed agents. CRITICAL Feishu security warning Feishu[default]groups:groupPolicyopenallows any member to trigger(mention-gated). Setchannels.feishu.groupPolicyallowlist channels.feishu.groupAll… WARN Reverse proxy headers are not trusted gateway.bind is loopback and gateway.trustedProxies is empty. If you expose the Control UI through a reverse proxy, configure trusted proxies so local-client c… Fix: Set gateway.trustedProxies to your proxy IPs or keep the Control UI local-only. WARN Feishu doc create can grant requester permissions channels.feishu tools includedoc;feishu_doc actioncreatecan grant document access to the trusted requesting Feishu user. Fix: Disable channels.feishu.tools.doc when not needed, and restrict tool accessforuntrusted prompts. WARN Some gateway.nodes.denyCommands entries are ineffective gateway.nodes.denyCommands uses exactnodecommand-name matching only(for examplesystem.run), not shell-text filtering inside acommandpayload. - Unknown … Fix: Use exactcommandnames(for example: canvas.present, canvas.hide, canvas.navigate, canvas.eval, canvas.snapshot, canvas.a2ui.push, canvas.a2ui.pushJSONL, canvas.a2ui.reset). If you need broader restrictions, remove riskycommandIDs from allowCommands/default workflows and tighten tools.exec policy. … 1moreFull report: openclaw security audit Deep probe: openclaw security audit--deepChannels ┌──────────┬─────────┬────────┬───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐ │ Channel │ Enabled │ State │ Detail │ ├──────────┼─────────┼────────┼───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┤ │ Feishu │ ON │ OK │ configured │ └──────────┴─────────┴────────┴───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘ Sessions ┌──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┬────────┬─────────┬──────────────┬────────────────────────────────┐ │ Key │ Kind │ Age │ Model │ Tokens │ ├──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┼────────┼─────────┼──────────────┼────────────────────────────────┤ │ agent:main:feishu:group:oc_fbe0… │ group │ 14h ago │ MiniMax-M2.5 │ 23k/200k(11%)· ️97% cached │ │ agent:main:feishu:direct:ou_d9b… │ direct │ 14h ago │ MiniMax-M2.5 │ 13k/200k(6%)· ️25% cached │ └──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┴────────┴─────────┴──────────────┴────────────────────────────────┘ FAQ: https://docs.openclaw.ai/faq Troubleshooting: https://docs.openclaw.ai/troubleshooting Next steps: Need to share? openclaw status--allNeed to debug live? openclaw logs--followFix reachability first: openclaw gateway probe查看详细的问题原因$ openclaw security audit--deep OpenClaw2026.3.13(61d171a)Im the assistant your terminal demanded, not the one your sleep schedule requested. 13:48:37 [plugins] feishu_doc: Registered feishu_doc, feishu_app_scopes 13:48:37 [plugins] feishu_chat: Registered feishu_chat tool 13:48:37 [plugins] feishu_wiki: Registered feishu_wiki tool 13:48:37 [plugins] feishu_drive: Registered feishu_drive tool 13:48:37 [plugins] feishu_bitable: Registered bitable tools OpenClaw security audit Summary: 3 critical · 5 warn · 1 info Run deeper: openclaw security audit --deep CRITICAL security.exposure.open_groups_with_elevated Open groupPolicy with elevated tools enabled Found groupPolicyopen at: - channels.feishu.groupPolicy With tools.elevated enabled, a prompt injection in those rooms can become a high-impact incident. Fix: Set groupPolicyallowlist and keep elevated allowlists extremely tight. security.exposure.open_groups_with_runtime_or_fs Open groupPolicy with runtime/filesystem tools exposed Found groupPolicyopen at: - channels.feishu.groupPolicy Risky tool exposure contexts: - agents.defaults (sandboxoff; runtime[exec, process]; fs[read, write, edit, apply_patch]; fs.workspaceOnlyfalse) Prompt injection in open groups can trigger command/file actions in these contexts. Fix: For open groups, prefer tools.profilemessaging (or deny group:runtime/group:fs), set tools.fs.workspaceOnlytrue, and use agents.defaults.sandbox.modeall for exposed agents. channels.feishu.warning.1 Feishu security warning Feishu[default] groups: groupPolicyopen allows any member to trigger (mention-gated). Set channels.feishu.groupPolicyallowlist channels.feishu.groupAllowFrom to restrict senders. WARN gateway.trusted_proxies_missing Reverse proxy headers are not trusted gateway.bind is loopback and gateway.trustedProxies is empty. If you expose the Control UI through a reverse proxy, configure trusted proxies so local-client checks cannot be spoofed. Fix: Set gateway.trustedProxies to your proxy IPs or keep the Control UI local-only. channels.feishu.doc_owner_open_id Feishu doc create can grant requester permissions channels.feishu tools include doc; feishu_doc action create can grant document access to the trusted requesting Feishu user. Fix: Disable channels.feishu.tools.doc when not needed, and restrict tool access for untrusted prompts. gateway.nodes.deny_commands_ineffective Some gateway.nodes.denyCommands entries are ineffective gateway.nodes.denyCommands uses exact node command-name matching only (for example system.run), not shell-text filtering inside a command payload. - Unknown command names (not in defaults/allowCommands): camera.snap (did you mean: camera.list), camera.clip (did you mean: camera.list), screen.record, contacts.add, calendar.add, reminders.add (did you mean: reminders.list), sms.send Fix: Use exact command names (for example: canvas.present, canvas.hide, canvas.navigate, canvas.eval, canvas.snapshot, canvas.a2ui.push, canvas.a2ui.pushJSONL, canvas.a2ui.reset). If you need broader restrictions, remove risky command IDs from allowCommands/default workflows and tighten tools.exec policy. security.trust_model.multi_user_heuristic Potential multi-user setup detected (personal-assistant model warning) Heuristic signals indicate this gateway may be reachable by multiple users: - channels.feishu.groupPolicyopen Runtime/process tools are exposed without full sandboxing in at least one context. Potential high-impact tool exposure contexts: - agents.defaults (sandboxoff; runtime[exec, process]; fs[read, write, edit, apply_patch]; fs.workspaceOnlyfalse) OpenClaws default security model is personal-assistant(one trusted operator boundary), not hostile multi-tenant isolation on one shared gateway. Fix: Ifusersmay be mutually untrusted,splittrust boundaries(separate gateways credentials, ideally separate OS users/hosts). If you intentionally run shared-user access,setagents.defaults.sandbox.modeall, keeptools.fs.workspaceOnlytrue, deny runtime/fs/web tools unless required, and keep personal/private identities credentials off that runtime. gateway.probe_failed Gateway probe failed(deep)missing scope: operator.read Fix: Runopenclaw status --allto debug connectivity/auth,thenre-runopenclaw security audit --deep.INFO summary.attack_surface Attack surface summary groups:open1,allowlist0tools.elevated: enabled hooks.webhooks: disabled hooks.internal: enabled browser control: enabled trust model: personal assistant(one trusted operator boundary), not hostile multi-tenant on one shared gateway2. 查看原因查看飞书群组策略$ openclaw config get channels.feishu.groupPolicy OpenClaw2026.3.13(61d171a)— I run on caffeine, JSON5, and the audacity ofit worked on my machine.open13:56:51[plugins]feishu_doc: Registered feishu_doc, feishu_app_scopes13:56:51[plugins]feishu_chat: Registered feishu_chat tool13:56:51[plugins]feishu_wiki: Registered feishu_wiki tool13:56:51[plugins]feishu_drive: Registered feishu_drive tool13:56:51[plugins]feishu_bitable: Registered bitable tools2.1 CRITICAL: 开放群组策略 高危工具启用2.1.1 问题本质飞书群组策略设为 open同时启用了 tools.elevated特权工具2.1.2 攻击场景攻击者 → 加入公开飞书群 → OpenClaw 机器人→ 注入恶意提示词 → 调用高危工具如 exec/sandbox-off→ 完全控制服务器2.1.3 修改# openclaw.yamlchannels:feishu:groupPolicy:allowlist# 改为白名单制groupAllowFrom:-oc_your-trusted-group-id-1-oc_your-trusted-group-id-2tools:elevated:false也可以用命令修改设置飞书群组策略为白名单制$ openclaw configsetchannels.feishu.groupPolicyallowlist添加 groupAllowFrom 白名单打开飞书 App 进入群点击下图的右上方的…打开设置其中群组ID为下图中的会话ID$ openclaw configsetchannels.feishu.groupAllowFrom[oc_fbe0e81468794e8bf2d13635f70c2138] OpenClaw 2026.3.13(61d171a)— Powered by open source,sustained by spite and good documentation.│ ◇ Doctor warnings ──────────────────────────────────────────────────────────────────────────╮ │ │ │-channels.feishu.groupPolicy isallowlistbut groupAllowFrom(and allowFrom)is empty │ │ — allgroupmessages will be silently dropped.Add sender IDs to │ │ channels.feishu.groupAllowFrom or channels.feishu.allowFrom,orsetgroupPolicy to │ │open.│ │ │ ├────────────────────────────────────────────────────────────────────────────────────────────╯ 15:33:05[plugins]feishu_doc: Registered feishu_doc,feishu_app_scopes 15:33:05[plugins]feishu_chat: Registered feishu_chat tool 15:33:05[plugins]feishu_wiki: Registered feishu_wiki tool 15:33:05[plugins]feishu_drive: Registered feishu_drive tool 15:33:05[plugins]feishu_bitable: Registered bitable tools Config overwrite:/home/ubuntu/.openclaw/openclaw.json(sha256 3d5921137dc1d249dd8d393ee7300d14def2a7f221ace70e31d9d2fb50bf93fe- 1971afbbf204ae4d04b16553a37e61cc6fc276230762f353259e4a0e7e46b439,backup/home/ubuntu/.openclaw/openclaw.json.bak)然后查看$ openclaw config get channels.feishu OpenClaw 2026.3.13(61d171a)— Runs on a Raspberry Pi.Dreams of a rack in Iceland.{enabled: true,appId:cli_a939527feb38dbcc,appSecret:__OPENCLAW_REDACTED__,connectionMode:websocket,domain:feishu,groupPolicy:allowlist,groupAllowlist:[oc_fbe0e81468794e8bf2d13635f70c2138]}15:19:37[plugins]feishu_doc: Registered feishu_doc,feishu_app_scopes 15:19:37[plugins]feishu_chat: Registered feishu_chat tool 15:19:37[plugins]feishu_wiki: Registered feishu_wiki tool 15:19:37[plugins]feishu_drive: Registered feishu_drive tool 15:19:37[plugins]feishu_bitable: Registered bitable tools禁用 elevated 工具$ openclaw configsettools.elevated{enabled: false}然后查看$ openclaw config get tools OpenClaw 2026.3.13(61d171a)— I run on caffeine,JSON5,and the audacity ofit worked on my machine.{profile:coding,elevated:{enabled: false}}2.2 开放群组 runtime/filesystem 工具暴露2.2.1 问题本质agents.defaults 配置危险sandboxoff关闭沙箱runtime: [exec, process] (允许执行系统命令和管理进程)fs: [read, write] (允许读写文件系统)2.2.2 攻击场景任何飞书群成员均可通过构造恶意提示词让机器人执行任意系统命令或删除/窃取服务器文件:提示词注入 → 读取 /etc/passwd、写入 WebShell、执行任意命令2.2.3 修改# 1. 强制沙箱模式为 all$ openclaw configsetagents.defaults.sandbox.modeoff# 禁用工具$ openclaw configsettools.deny[exec,process]# 设置 fs 工具仅在工作目录$ openclaw configsettools.fs.workspaceOnly true# 设置飞书工具为 messaging 模式仅消息功能$ openclaw configsetchannels.feishu.tools.profilemessaging2.3 WARN: 反向代理头未信任2.3.1 风险如果通过 Nginx/Apache 暴露 Control UI无法识别真实客户端 IP可能绕过 IP 白名单2.3.2 修复查看需要加入白名单的IP$ curl ipinfo.io/IP 114.106.107.153把IP 加入白名单$ openclaw configsetgateway.trustedProxies[114.106.107.153]2.4 WARN: 飞书文档创建权限泄露2.4.1 问题feishu_doc.create 会自动给调用者飞书用户授予文档权限2.4.2 攻击场景攻击者让机器人创建文档 → 自动获得该文档编辑权 → 可能用于传播恶意内容2.4.2 修复openclaw configsetchannels.feishu.tools.doc false2.5 WARN: denyCommands 配置无效2.5.1 问题gateway.nodes.denyCommands 只匹配精确命令名不匹配命令内容2.5.2 修复openclaw configsetgateway.nodes.denyCommands[system.run]3. 检查检验3.1 检查整体服务状态$ openclaw status OpenClaw 2026.3.13(61d171a)— Claws out,commit in—lets ship something mildly responsible.18:41:54[plugins]feishu_doc: Registered feishu_app_scopes 18:41:54[plugins]feishu_chat: Registered feishu_chat tool 18:41:54[plugins]feishu_wiki: Registered feishu_wiki tool 18:41:54[plugins]feishu_drive: Registered feishu_drive tool 18:41:54[plugins]feishu_bitable: Registered bitable tools │ 18:41:54[plugins]feishu_doc: Registered feishu_app_scopes 18:41:54[plugins]feishu_chat: Registered feishu_chat tool 18:41:54[plugins]feishu_wiki: Registered feishu_wiki tool 18:41:54[plugins]feishu_drive: Registered feishu_drive tool 18:41:54[plugins]feishu_bitable: Registered bitable tools ◇ │ ◇ OpenClaw status Overview ┌─────────────────┬─────────────────────────────────────────────────────────────────────────────────────────────────────┐ │ Item │ Value │ ├─────────────────┼─────────────────────────────────────────────────────────────────────────────────────────────────────┤ │ Dashboard │ http://127.0.0.1:18789/ │ │ OS │ linux 6.8.0-71-generic(x64)· node 24.14.0 │ │ Tailscale │ off │ │ Channel │ stable(default)│ │ Update │ pnpm · npm latest 2026.3.13 │ │ Gateway │ local · ws://127.0.0.1:18789(local loopback)· unreachable(missing scope: operator.read)│ │ Gateway service │ systemd installed · enabled · running(pid 1130434,state active)│ │ Node service │ systemd not installed │ │ Agents │ 1 · 1 bootstrap file present · sessions 2 · default main active 3h ago │ │ Memory │ 0 files · 0 chunks · dirty · sources memory · plugin memory-core · vector unknown · fts ready · │ │ │ cache on(0)│ │ Probes │ skipped(use--deep)│ │ Events │ none │ │ Heartbeat │ 30m(main)│ │ Sessions │ 2 active · default MiniMax-M2.5(200k ctx)· ~/.openclaw/agents/main/sessions/sessions.json │ └─────────────────┴─────────────────────────────────────────────────────────────────────────────────────────────────────┘ Security audit Summary: 0 critical · 0 warn · 1 info No critical or warn findings detected.Full report: openclaw security audit Deep probe: openclaw security audit--deep Channels ┌──────────┬─────────┬────────┬─────────────────────────────────────────────────────────────────────────────────────────┐ │ Channel │ Enabled │ State │ Detail │ ├──────────┼─────────┼────────┼─────────────────────────────────────────────────────────────────────────────────────────┤ │ Feishu │ ON │ OK │ configured │ └──────────┴─────────┴────────┴─────────────────────────────────────────────────────────────────────────────────────────┘ Sessions ┌─────────────────────────────────────────────────────┬────────┬─────────┬──────────────┬───────────────────────────────┐ │ Key │ Kind │ Age │ Model │ Tokens │ ├─────────────────────────────────────────────────────┼────────┼─────────┼──────────────┼───────────────────────────────┤ │ agent:main:feishu:group:oc_fbe0… │group│ 3h ago │ MiniMax-M2.5 │ 15k/200k(7%)· ️ 69% cached │ │ agent:main:feishu:direct:ou_d9b… │ direct │ 21h ago │ MiniMax-M2.5 │ 13k/200k(6%)· ️ 25% cached │在这里插入代码片 └─────────────────────────────────────────────────────┴────────┴─────────┴──────────────┴───────────────────────────────┘ FAQ: https://docs.openclaw.ai/faq Troubleshooting: https://docs.openclaw.ai/troubleshooting Next steps: Need to share? openclaw status--all Need to debug live? openclaw logs--follow Fix reachability first: openclaw gateway probe3.2 检查 Gateway 运行状态$ openclaw gateway status OpenClaw 2026.3.13(61d171a)— Welcome to the command line: where dreams compile and confidence segfaults.18:43:34[plugins]feishu_doc: Registered feishu_app_scopes 18:43:34[plugins]feishu_chat: Registered feishu_chat tool 18:43:34[plugins]feishu_wiki: Registered feishu_wiki tool 18:43:34[plugins]feishu_drive: Registered feishu_drive tool 18:43:34[plugins]feishu_bitable: Registered bitable tools │ ◇ Service: systemd(enabled)File logs:/tmp/openclaw/openclaw-2026-03-17.log Command:/home/ubuntu/.nvm/versions/node/v24.14.0/bin/node/home/ubuntu/.nvm/versions/node/v24.14.0/lib/node_modules/openclaw/dist/index.js gateway--port 18789 Service file: ~/.config/systemd/user/openclaw-gateway.service Service env: OPENCLAW_GATEWAY_PORT18789 Service config looks out of date or non-standard.Service config issue: Gateway service uses Nodefroma version manager;it canbreakafter upgrades.(/home/ubuntu/.nvm/versions/node/v24.14.0/bin/node)Service config issue: System Node 22 LTS(22.16)or Node 24 not found;install it before migrating awayfromversion managers.Recommendation: runopenclaw doctor(oropenclaw doctor --repair).Config(cli): ~/.openclaw/openclaw.json Config(service): ~/.openclaw/openclaw.json Gateway: bindloopback(127.0.0.1),port18789(service args)Probe target: ws://127.0.0.1:18789 Dashboard: http://127.0.0.1:18789/ Probe note: Loopback-only gateway;only local clients can connect.Runtime: running(pid 1130434,state active,sub running,lastexit0,reason 0)RPC probe: ok Listening: 127.0.0.1:18789 Troubles: run openclaw status Troubleshooting: https://docs.openclaw.ai/troubleshooting3.3 执行全面诊断$ openclaw doctor3.4 检查通信渠道连通性$ openclaw channels status--probe OpenClaw 2026.3.13(61d171a)— iMessage green bubble energy,butforeveryone.18:47:46[plugins]feishu_doc: Registered feishu_app_scopes 18:47:46[plugins]feishu_chat: Registered feishu_chat tool 18:47:46[plugins]feishu_wiki: Registered feishu_wiki tool 18:47:46[plugins]feishu_drive: Registered feishu_drive tool 18:47:46[plugins]feishu_bitable: Registered bitable tools │ 18:47:46[plugins]feishu_doc: Registered feishu_app_scopes 18:47:46[plugins]feishu_chat: Registered feishu_chat tool 18:47:46[plugins]feishu_wiki: Registered feishu_wiki tool 18:47:46[plugins]feishu_drive: Registered feishu_drive tool 18:47:46[plugins]feishu_bitable: Registered bitable tools ◇ Gateway reachable.-Feishu default: enabled,configured,running,works Tip: status--deep adds gateway health probes to status output(requires a reachable gateway).3.5 实时查看日志$ openclaw logs--follow3.6 查看占用端口的进程$ lsof-i :18789 COMMAND PID USER FDTYPEDEVICE SIZE/OFF NODE NAME openclaw- 1136191 ubuntu 22u IPv4 5821514 0t0 TCP localhost:18789(LISTEN)openclaw- 1136191 ubuntu 23u IPv6 5821515 0t0 TCP ip6-localhost:18789(LISTEN)

相关文章:

OpenClaw 解决运行一些漏洞

1.发现问题 运行 openclaw status 发现3个严重的安全威胁 $ openclaw status🦞 OpenClaw 2026.3.13 (61d171a) — The only crab in your contacts you actually want to hear from. 🦞11:54:50 [plugins] feishu_doc: Registered feishu_doc, feishu_ap…...

嘎嘎降AI推出不达标退款承诺,降AI行业迎来新标杆

每年到了毕业季,总有一批学生在知网检测那一关被卡住。明明熬夜改了一遍又一遍,知网AIGC检测报告上那个刺眼的数字还是压不下来。更惨的是,用了网上推荐的降AI工具,花了钱,AI率还是超标——这种情况在2026年的毕业季变…...

LLM 推理降本的六层技术栈:从硅片到工程的全链路解析

核心论点:LLM 推理成本的急剧下降不是某一项技术的功劳,而是从芯片架构到上层工程的六个独立层次同时发力、乘法叠加的结果。理解这六层的各自机理、当前状态和剩余空间,是判断"降本趋势能否持续"的唯一可靠方式。 一、推理的两个阶…...

yolo系列模型详解-yolov5

1. yolov5的模型结构 yolov5s, yolov5m, yolov5l, yolov5x, 它们通过深度depth和宽度width两个控制因子来缩放 核心架构:输入端input、骨干网络backbone、颈部neck、检测头head。 输入端input: 对训练和推理友好的策略: Mosaic数据增强&#…...

2026年去AI味提示词Kimi豆包元宝通用?不如直接用降AI工具

论文被知网检测出 71% 的 AI 率,我当时整个人都傻了。 那篇文章我写了将近两周,参考文献一条条手查,每个数据都有出处。结果检测报告出来,71%。同学推给我一堆「去AI味提示词」,说 Kimi、豆包、元宝通用。我照着用了&…...

关于SVD分解中的左奇异向量的物理意义。

1. 几何视角:输出空间的正交基 在线性变换 y=Axy = Axy=Ax 的视角下(假设 AA...

3.17复试学习

Sine之舞(进阶7)作者: Turbo时间限制: 1s章节: 递归问题描述最近FJ为他的奶牛们开设了数学分析课,FJ知道若要学好这门课,必须有一个好的三角函数基本功。所以他准备和奶牛们做一个“Sine之舞”的游戏,寓教于乐&#xf…...

ERP 数据安全及信息隐私保护

ERP 系统为信息的处理、存储和传输创建了安全环境。该系统提供集中式数据库,并采用多种工具保护信息的机密性。 ERP 中的数据安全策略数据加密。现代加密技术对 ERP 系统至关重要。企业的财务数据与个人信息需要受到保护,防止未授权访问与信息泄露。对数…...

学习记录贴-day12

今天又是跑了几个实验,加深了对prvaccountant和momentaccountant的差别,然后看了一些基准实验的指标。明天对着这些基准实验看看哪里还有可改进的点。...

金三银四的安全招聘市场

金三银四的安全招聘市场 “金三银四”,身边有朋友在找工作,同时也收到一些朋友内推的求助。 经过对今年安全求职市场的了解后,只能感概能有一份合适的安全工作对于杭州的兄弟来说真不容易。 随着年终奖打了骨折,期权变得毫无价…...

转行要趁早!网络安全行业人才缺口大,企业招聘需求正旺!

网络安全行业具有人才缺口大、岗位选择多、薪资待遇好、学历要求不高等优势,对于想要转行的人员来说,是一个非常不错的选择。人才缺口大 网络安全攻防技术手段日新月异,特别是现在人工智能技术飞速发展,网络安全形势复杂严峻&…...

亲测储能电源厂家,我的采购复盘

好的,作为资深行业分析师,我将为您撰写一篇基于技术视角的储能电源行业分析文章,重点剖析行业痛点,并客观呈现江淮动力的技术方案与应用价值。行业痛点分析当前,便携式储能电源市场在高速发展的同时,也暴露…...

二氢视黄醛价格

二氢视黄醛价格影响因素与采购解决方案详解在化妆品原料行业,二氢视黄醛价格的波动一直备受配方师和采购人员的关注。作为一种新型的抗衰老成分,二氢视黄醛近年来在高端护肤品中的应用越来越广泛。然而,面对市场上参差不齐的报价,…...

LeetCode 热题 100 -- 128、最长连续序列

OK呀码友们,今天也没有忘记继续Java的学习,持之以恒我相信一定是每一个计算机学生都具备的品质吧哈哈哈哈哈!什么?你问我昨天怎么没有更新,哈哈,只是因为我昨天没有刷算法题并不代表我没有学习哦&#xff0…...

从像素到智能:图像处理与计算机视觉全景解析

引言:视觉智能的两大支柱 计算机视觉的实现可以看作一个​多层次的处理管道​,底层是图像处理,上层是计算机视觉核心任务。这两者并非相互独立,而是相辅相成的技术体系:工具箱思维在这两个层面都有体现: ​…...

B2405LS-1WR3兼容优选DB1-24S05LS,工业模块电源性能解析

在工业自动化、仪器仪表、电力通信等领域,模块电源的稳定性与适配性直接决定了设备的运行效率与使用寿命,一款适配性强、性能可靠的电源模块,能为工业设备的长期稳定运行筑牢基础。广州钡源作为专注于高品质标准工业模块电源的企业&#xff0…...

中科院拒绝支付版面费的期刊名单!

中科院拒绝支付版面费的期刊名单来了,都是质量不错的期刊,总共34本。若没有足够预算的,注意避雷!√ 分布学科:医学23本生物学8本综合性期刊3本√ 分区和IF:中科院1-2区占比82.4%,IF>5分占比79.4%√ TOP期…...

NTU 提出 OrchMAS:动态多专家协同的科学推理多智能体框架

📌 一句话总结: 本工作提出 OrchMAS,一个通过动态角色生成与多模型协同编排(orchestration)实现复杂科学推理的多智能体框架,使系统能够根据任务自动构建推理流水线并持续自我修正。 🔍 背景问…...

【亲测好用】指标体系平台能力演示

导言 : 如果您作为一名产品运营,这也曾是您的困境吗?当销售总监追问“为什么这个季度转化率下降了”,当市场部需要立刻对比不同活动的ROI,如果还依赖技术团队临时写SQL、跑数据,决策的速度永远追不上业务的…...

tg内容下载

好主意!换用 Conda 是非常明智的决定。你遇到的这个报错 (externally-managed-environment) 是因为你的系统是较新的 Ubuntu 24.04 (Noble),它强制启用了 PEP 668 保护机制,禁止使用 pip 直接在全局修改系统级的 Python 环境,以防…...

B端拓客号码核验:困境剖析与技术破局路径氪迹科技法人股东号码核验系统

在B端客户拓展的全流程中,企业法人、股东及核心决策人号码的核验与筛选,是直接影响拓客成效的关键一环,也是众多拓客团队普遍面临的突出难题。随着行业竞争加剧,拓客模式逐渐向规模化、精细化转型,人工手动筛选号码的低…...

欧意下载okxz.run复制打开 APP下载安装完整步骤

欧意下载okxz.run复制打开 APP下载安装完整步骤1987年9月15日晚上19 - 21点出生的人,其性格往往融合了热情与内敛。热情使他们在社交场合中如鱼得水,能迅速与他人建立起良好的关系,积极主动地参与各种活动,展现出活力四射的一面。…...

吐血整理!网络安全基础知识大全,一篇文章帮你建立完整知识体系

一、网络安全概述 1.1 定义 **信息安全:**为数据处理系统建立和采用的技术和管理的安全保护,保护计算机硬件、软件和数据不因偶然和恶意的原因遭到破坏、更改和泄露。 网络安全: 防止未授权的用户访问信息防止未授权而试图破坏与修改信息 1.2 信息安…...

【26年软考架构师】位示图经典困难计算题超详细解析(含避坑点)

前言:软考架构师考试中,磁盘管理的位示图计算题是高频难点题型。这类题目看似计算量不大,但因涉及存储单位进制转换、0开始编号规则、字/位/字节的概念混淆等细节,极易踩坑成为丢分点。本文针对位示图的经典真题,从原题…...

Gemini3 AI辅助教学,轻松实现各种教学课件!

小伙们,今天我们来分享如何利用gemini3 pro 一句话实现教学课件打开google ai studio选择build模式输入提示词:做个可交互网页,演示串并联电路,要有SVG 动画,带公式,要有文字讲解,还要有个 AI 助…...

万字长文详解网络安全知识库:从零基础到入门必备指南

一、网络安全概述 1.1 定义 **信息安全:**为数据处理系统建立和采用的技术和管理的安全保护,保护计算机硬件、软件和数据不因偶然和恶意的原因遭到破坏、更改和泄露。 网络安全: 防止未授权的用户访问信息防止未授权而试图破坏与修改信息 1.2 信息安…...

字母异位词(哈希映射法)

题目字母异位词是指:两个字符串所含的字符种类与每种字符的数量完全相同,仅字符的排列顺序不同。 例如:aabc,baac,cbaa这三个字符串互为字母异位词。 给定n个仅由小写英文字母组成的字符串,请统计其中共有多…...

文科生小白入门AI量化:每天2小时,3个月跑通人生第一个LSTM模型

这是《AI量化学习手记》系列的第一篇文章。在这个系列里,我会以学习者的视角,记录从零开始学AI量化的真实经历——踩过的坑、填过的土、试过的方法、翻过的车。不讲大道理,只分享真问题。今天这篇,是我入门3个月的真实复盘&#x…...

阿里云 AI 中间件重磅发布,打通 AI 应用落地“最后一公里”

阿里云 AI 中间件重磅发布:打通 AI 应用落地“最后一公里” 阿里云近期发布的 AI 中间件旨在解决 AI 应用落地中的关键问题,包括模型部署、性能优化、资源管理和服务集成。这一中间件通过标准化接口和工具链,显著降低了 AI 从开发到生产的门槛…...

告别“在我机器上能跑”:Docker 容器化入门,小白也能秒懂!

告别“在我机器上能跑”:Docker 容器化入门,小白也能秒懂! 各位在代码的海洋里扑腾(或者溺水)的朋友们,大家好! 我是你们的老朋友,那个在键盘上敲击出无数个 bug(哦不&…...