Sign-Off

在每次完整输出后追加个性化落款签名。像对讲机的 Over，让用户明确知道 AI 说完了。Use when completing any response or after tool calls return results.

Security Scan

The skill's files, instructions, and behavior are consistent with its stated purpose (appending a rendered signature at the end of outputs); it does not request credentials or perform network access in the provided code.

Purpose & Capability

Name/description (append a signature) matches the delivered artifacts: SKILL.md describes reading a sign-off.json and templates/, and render.js plus template files implement variable rendering. No unrelated binaries, credentials, or config paths are requested.

Instruction Scope

Runtime instructions constrain the skill to reading sign-off.json in the workspace and rendering templates; render.js only reads local files and prints the rendered signature. SKILL.md mentions optional weather API usage as a possibility, but the included code does not call external APIs. The skill also describes updating sign-off.json via conversational commands — that implies the agent (or user) may write to a file in the workspace when configured to do so.

Install Mechanism

No install spec; this is instruction-only with source files bundled. There are no external downloads, package installs, or extract steps. package.json exists but has no dependencies; render.js is local and has no network calls.

Credentials

The skill requires no environment variables or credentials. The only environment read by the code is OPENCLAW_WORKSPACE (optional) to locate the workspace root — this is proportional to locating sign-off.json.

Persistence & Privilege

always is false and the skill does not request system-wide privileges. However, its operation expects access to the workspace root (reading sign-off.json and templates) and SKILL.md suggests updating sign-off.json via conversation — so installing/using it may result in writes to your workspace config if you accept or issue those changes.

Assessment

This skill appears coherent and low-risk, but review a few things before installing: - Inspect sign-off.json (workspace root) and templates so you know what will be read and what a rendered signature might contain. - Be aware that telling the agent to "change your signature" implies writing to sign-off.json in your workspace; grant write access only if you trust the agent and the skill. - The included render.js reads local files only and does not call external services; if you (or a future change to the skill) add a weather API or other network calls, review those endpoints and any required API keys. - If installing from the suggested Git repo, verify the repository source/author (git clone URL in README) if provenance matters to you. No credentials or special system privileges are requested by the skill as provided.

Sign-Off — AI 输出结束标记

在每次完整输出的末尾追加一个个性化落款，让用户明确知道你说完了。

触发时机

在以下情况的最后追加签名：

• 完成一次完整回复后
• 工具调用返回结果并给出总结后
• 多段输出的最后一段

不触发

• 还在思考/输出中（不要在中间段落加）
• 只是状态更新（如"正在处理..."）
• NO_REPLY 场景

签名格式

另起一行，在回复内容的最末尾追加：

回复内容...

      {签名内容}

签名前加缩进（4空格或1个Tab），视觉上与正文区分。

读取配置

每次输出前，检查 workspace 根目录下是否有 sign-off.json：

有配置文件： 读取模板和变量，渲染签名 无配置文件： 使用默认签名 {agentName} · {location}

变量渲染

替换模板中的变量。对于自动变量（season、weather、time 等），可使用 render.js 脚本自动填充：

node render.js              # 默认模板
node render.js --work       # 使用 work 模板
node render.js --casual     # 使用 casual 模板

变量列表：

变量	来源	示例
{name}	sign-off.json → name	小橘
{location}	sign-off.json → location	杭州
{emoji}	sign-off.json → emoji	🏮
{seal}	sign-off.json → seal	[橘印]
{season}	自动判断	春/夏/秋/冬
{weather}	自动（可查天气API）	晴/雨
{time}	自动（当前时段）	午后/深夜/清晨
{greeting}	自动（时段问候）	早安/午安/夜安
{dayOfWeek}	自动	周五
{zodiac}	自动	蛇年
{mood}	sign-off.json → mood	慵懒

上下文模式

如果配置了 contextMode，根据当前对话场景选择模板：

• work：技术讨论、执行命令 → 使用 work 模板
• casual：闲聊、回答问题 → 使用 casual 模板
• auto（默认）：AI 自行判断

如果 contextMode 未配置，统一使用 template 字段。

口头配置

用户可以口头修改签名：

"你的签名改成 From 小橘, Hangzhou, with 🧡"

收到此类指令时，更新 sign-off.json 中的对应字段。

模板安装

用户也可以从模板库安装预设风格：

"给我换个书法风的签名"

此时读取 templates/ 目录下对应的 JSON 文件，合并到 sign-off.json。