跳到正文
cccskills
免费在 GitHub 公开

amq-remote

Let the owner reach this running agent from Buzz, on the phone or in Buzz Desktop. Use when the user types /amq-remote (Claude Code) or $amq-remote (Codex), /amq-remote --native, /amq-remote off, or /amq-remote status, or asks to control this session from Buzz. Not for AMQ messaging between agents (use amq-cli).

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md7.8 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

AMQ Remote

/amq-remote gives this session its own Buzz agent, with its own DM. The owner DMs that agent from the Buzz phone app or Buzz Desktop. Several sessions can be connected at once; each has its own agent. The same session connecting again gets its own agent back. A different session never gets another session's agent.

In Codex the skill is invoked as $amq-remote (Codex does not register / commands for skills); the commands below are the same. Codex runs sandboxed: attach writes ~/.amq/remote, so the first attach may be refused until the user reruns it with approval — ask for that one approval, do not work around it.

  • Default (AMQ mailbox). Each DM arrives as an AMQ message from buzz in this agent's inbox. This agent answers with amq reply --id <id>, and that reply is shown in the DM. A reply with --kind status shows as progress; any other reply is the final answer. No Stop hook and no wake are needed. Noticing the message is this agent's own business (a wake, a built-in consumer, monitor, or the next drain).
  • --native. For a session that is not on AMQ: each DM is typed into this exact conversation, and the harness's own final answer goes back. On Claude Code it needs the AMQ Stop hook.

Run the steps in order. When a step says to stop, show the user the command's own message. Do not work around a refusal.

/amq-remote (connect)

  1. Tools. Run command -v amq-remote amq-acp. If either is missing, tell the user to run brew install avivsinai/tap/amq, then stop.
  2. Connect.
    • Default: run amq-remote attach --self --json. It uses AM_ROOT and AM_ME. If it says this session is not an AMQ participant, offer --native.
    • --native: use $AM_ROOT when set, else $HOME/.amq/remote/root (create it private with (umask 077 && mkdir -p "$HOME/.amq/remote/root")). Run amq-remote attach --self --native --root "$ROOT" --json.
    • If it exits 6 and names amq-remote claude install-stop-hook (Claude Code with --native), tell the user this adds one Stop hook to ~/.claude/settings.json; it always exits 0, and amq-remote claude uninstall-stop-hook removes it. Ask once, run it, then run the attach again.
    • Any other failure: show its message and stop.
  3. Tell the owner only what is still needed. The JSON has name, label, agent, helper, draft, reused, and agent_name when there is an agent. Say "Connected: <label>." and then exactly one of:
    • agent is running: "DM <agent_name> in Buzz." Nothing else is needed.
    • agent is offline: "Start <agent_name> in Buzz Desktop (Agents, the agent, Start), then DM it. If you deleted it, run amq-acp setup --session <name> and import the file."
    • agent is missing and draft is true: "Buzz Desktop opens Create agent for it: pick AMQ Remote, then Create and start."
    • agent is missing and helper is missing: this machine has no helper yet. Run amq-acp setup --helper --out "$HOME/Downloads/AMQ Remote helper.agent.json" once, and show its printed steps as they are (Buzz Desktop: Agents, then + then Import, pick that file, then Start, and turn on Auto-start). Later sessions then need only the Create agent dialog.
    • agent is missing otherwise: this session's one-time setup. Run amq-acp setup --session <name> --out "$HOME/Downloads/AMQ <name>.agent.json" and show its printed steps as they are. If the JSON has warning, show it too.
  4. --native on Claude Code only: the approval hook. Run amq-remote doctor --root "$ROOT" --json and read its failing list; the exit code is not the signal. If a native_capability entry says the PermissionRequest hook is not installed and there is no relay share (the usual --native case), tell the user: "This adds one PermissionRequest hook. It exits at once in sessions that are not connected. It lets Buzz deny a tool call only: ❌ in the DM denies, and allow stays in Claude's terminal. amq-remote claude uninstall-approval-hook removes it." Ask once, then run amq-remote claude install-approval-hook. If the entry is for a Claude target shared through a relay share, tell the user: "This adds one PermissionRequest hook. It exits at once in sessions that are not shared. It pins your Buzz public key and this share's relay, body, DM channel and target, so only your signed ✅, or your yes in the approval's thread, in this DM can allow a tool call. Do not allow Claude to edit .claude for a session, use bypassPermissions mode, or allow every Bash command: each lets the pin be changed. The relay is trusted to return the message's full edit and deletion history; replacing the amq-remote binary, or a settings edit you approve, defeats the pin. Run the install again if the share's relay or channel changes. Another decision hook you have can answer first. amq-remote claude uninstall-approval-hook removes it." Ask once, then run amq-remote claude install-approval-hook --root "$ROOT". If doctor reports claude_approval_pin_warnings, show them to the user.
  5. Add one line for the carrier:
    • Default: "DMs arrive in this agent's AMQ inbox. Stop in Buzz stops waiting; a delivered message may still be acted on."
    • --native: "A DM sent while this session is busy waits. Buzz Stop cannot interrupt a Claude turn."
    • When the approval hook was installed in step 4: "When a Buzz request needs a tool approval, the DM shows it. ❌ blocks that call. ✅, or yes replied in the approval's thread, allows it when the DM offers ✅: a Bash command shown whole. Otherwise allow it in the terminal. The first answer wins." Without a relay share, say instead: "When a Buzz request needs a tool approval, the DM shows it. ❌ denies that call; allow it in Claude's terminal. The first answer wins."

Codex with --native: attach needs CODEX_THREAD_ID and a thread loaded in the Codex app-server daemon. A session started with plain codex is not in the daemon; start it with codex app-server daemon start, then codex --remote unix://. Run amq-remote and amq-acp outside Codex's sandbox (request escalation): the sandbox blocks the daemon socket, and amq-acp setup writes files outside the workspace.

Answering a Buzz DM (default mailbox)

A message from buzz with subject "Buzz DM" is the owner writing from Buzz. Do the work, then answer with amq reply --id <that message id>. Use --kind status for an interim update. Keep the final answer short; it is read in a chat.

/amq-remote off

Run amq-remote detach --self. The session keeps running; its Buzz agent archives itself and its DM answers "Not connected". When this session connects again, the same agent comes back.

/amq-remote status

Run amq-remote bindings and show its output. It lists each connected session as live or unknown with its agent (running, offline or missing), prunes the sessions proven ended, and names, once, old agents the owner should delete.

Trust limits

The Buzz agent's identity belongs to Buzz Desktop. Its grant has no kind limit and no expiry. Archiving the agent does not invalidate a copied key and grant; only removing the agent's relay access does that. Buzz "owner only" also admits the owner's other agents. See the amq-acp README.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Use AtomLane to compile and execute safe atomic parallel plans on macOS and native Windows Preview for worthwhile independent argv tasks, dependency DAGs, supported platform entrypoints, or Apple-silicon operators. Use at task start or an execution boundary when structured local work may contain two or more worthwhile units; skip plain answers, one quick command, and work whose effects cannot be safely bounded.

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

add

免费

Register a deferred decision in the debt registry. Trigger by judgment, not a marker scan, whenever a future reader would ask "why this way?": an unmade decision, stub, loosened type, bypassed check, swallowed error, a default picked "for now", or a TODO/FIXME/HACK/XXX marker. Trigger immediately whenever you defer work, or when the user invokes $add. Over-register freely; the developer drops with "drop A", "drop A,C", or "drop all".

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

ADK 框架适配层。为 LangChain / EINO / AutoGen / AgentScope / CrewAI 提供框架特定的 代码模板、惯用模式、API 映射和项目结构,供 agent-dev-workshop Phase 5 代码生成使用。 每个框架 reference 文件标注 verified_date 用于版本锁定。

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

中文调试修复技能。用于报错、测试失败、页面异常、功能不符合预期、需要定位根因并做最小修复时。触发语包括"进入调试模式""帮我修问题""报错了""测试失败""页面坏了""找根因"。

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

交互式 AI Agent 开发工作坊:通过 6 阶段深度协作对话,引导用户完成 Agent 需求分析、架构设计、 工具定义、Prompt 与编排设计、代码生成、验证迭代,产出可直接运行的 Agent 项目。 框架无关设计优先,支持 LangChain / EINO / AutoGen / AgentScope / CrewAI 等 ADK 框架。

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

中文漂移审计技能。用于项目或学习过程变乱、上下文漂移、任务分叉、多个方案冲突、命名不一致、Codex 可能顺手改多了时。触发语包括"漂移检查""感觉跑偏了""项目变乱了""检查是否失控""分叉太多""上下文漂移"。

中文简介正在准备中。当前显示原文说明。

hashgraph-online/awesome-codex-plugins13502026年10月19日 更新

查看 hashgraph-online 的全部技能

报告此技能的问题