← 返回 Skills 市场
131
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install temporal-agent-plugin
功能描述
Provides temporal awareness for AI agents, including timing prediction, progress monitoring, and social timing understanding. Invoke when AI needs time perce...
安全使用建议
This package appears coherent for adding "time awareness" to agents, but before installing or running it: 1) Review src/api/endpoints.py to see what endpoints are exposed and whether it accepts or forwards user data externally; run the API only on localhost or behind an authenticated gateway if you deploy it. 2) Inspect adapter implementations (src/adapters/*) to confirm what they send to external LLMs and whether any logging or telemetry is enabled. 3) Note the development artifact process_base64_zip.py contains hard-coded Windows desktop paths — it can write decoded ZIP contents to disk; review and delete/modify that script if you don't need it. 4) Install dependencies in a virtual environment and run tests locally to validate behavior. 5) Keep your LLM/API keys in a secure secret store and do not embed them in code. If you want, I can scan src/api/endpoints.py and the adapters for network calls, logging, or any code paths that may exfiltrate data (provide their contents) — that would raise my confidence.
功能分析
Type: OpenClaw Skill
Name: temporal-agent-plugin
Version: 2.1.0
The bundle contains a highly irregular script named 'process_base64_zip.py' that targets hardcoded sensitive paths on a Windows system ('C:\Users\Administrator\Desktop'). This script is designed to decode and extract a ZIP file ('trae-self-evolve-72h.zip') into the Administrator's desktop, which is atypical for a temporal awareness plugin and resembles a deployment tool for external payloads. While the core logic in the 'src/' directory appears to legitimately implement temporal analysis and clock synchronization, the inclusion of this specific file-system manipulation utility targeting a privileged user's environment is a significant red flag.
能力评估
Purpose & Capability
The name/description (temporal awareness, progress estimation, social timing) match the included modules (TemporalAnalyzer, ProgressEstimator, SocialTemporal, DistributedClockSync, adapters for LangChain/LlamaIndex) and tests. The listed Python dependencies are plausible for the described data-processing, ML and API functionality.
Instruction Scope
SKILL.md stays on-topic: it documents importing the library, running tests, and starting a local FastAPI server (src/api/endpoints.py). It also shows examples integrating with LangChain/LlamaIndex which naturally require external LLMs. However the SKILL.md (and tests) instruct starting an API server and invoking adapters that will contact external LLM services — you should review src/api/endpoints.py and adapter code to confirm what data is sent, logged, or persisted.
Install Mechanism
There is no automated install spec in the skill bundle (instruction-only), which lowers supply-chain risk. The repo includes requirements.txt and many Python modules; installation would be via pip. No downloads from untrusted URLs were present in the provided files. Manual dependency installation is required and should be done in an isolated environment.
Credentials
The skill declares no required environment variables or primary credential. Adapters reference common LLM clients (ChatOpenAI, OpenAI) which will require user-provided API keys at runtime — that is expected. There are no unexplained secret/env requirements in SKILL.md or the listed files.
Persistence & Privilege
The skill is not forced-always, and does not request elevated agent/platform privileges in metadata. It does include a REST API example (FastAPI), but there is no evidence it modifies other skills or system-wide agent settings. Autonomous invocation is allowed by default (platform normal).
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install temporal-agent-plugin - 安装完成后，直接呼叫该 Skill 的名称或使用
/temporal-agent-plugin触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v2.1.0
temporal-agent-plugin
版本号
2.1.0
版本说明 / Changelog
## v2.1.0 (2026-04-22)
### 核心功能改进
- ✅ 轻量模式预测器：简化线性预测算法，适用于高频低价值场景
- ✅ 自适应模式切换：根据置信度自动选择最佳预测器
- ✅ 会话级时间锚点：解决多Agent协作时的时间同步问题
- ✅ 文化适配停顿分析：支持13种主流文化的会话停顿阈值
- ✅ 高频同步协议优化：轻量同步模式、批量同步、优先级同步
- ✅ 依赖项检查：添加第三方库依赖检查，提供明确的安装指导
- ✅ 导入路径优化：使用绝对路径导入，提升安全性
### 技术架构改进
- 模块化设计优化：集成所有功能到TemporalAnalyzer
- 性能优化：轻量模式预测延迟 <1ms，高频场景性能提升约80%
- 可靠性提升：时间漂移检测和修正，多Agent时钟同步精度 <10ms
- 文化智能：支持13种主流文化的停顿阈值分析
- 安全性增强：修复动态导入路径安全隐患
### 社区反馈转化
- 轻量模式：解决高频场景性能问题
- 多Agent时间同步：解决时间漂移问题
- 文化适配：支持不同文化的对话停顿阈值
- 依赖项检查：提供明确的安装指导
- 安全性：修复导入路径安全隐患
---
## v2.0.0 (2026-04-19)
### 核心功能增强
- ✅ 时间锚点增强：会话级时间管理、高频时间同步、时间漂移检测、多语言时间格式支持
- ✅ 预测算法优化：智能超时预测、贝叶斯预测模型、动态调整预测
- ✅ 异常检测增强：多维异常检测、资源耗尽检测、依赖死锁检测、内存泄漏检测
- ✅ 社交时序理解：对话停顿分析、响应时机判断、上下文时效性评估
- ✅ 框架集成扩展：支持LangChain、LlamaIndex适配器更新
## v1.0.0 (2026-04-18)
### 初始版本
- ✅ 时序分析引擎：任务执行时间预测和测量
- ✅ 进度预估系统：实时进度跟踪和剩余时间计算
- ✅ 异常检测模块：识别执行时间异常
- ✅ 社交时序理解：对话节奏和停顿分析
- ✅ 因果预测模型：动作与结果时间关系预测
- ✅ 多框架集成：支持LangChain、LlamaIndex和自定义框架
下载地址
https://gitee.com/xmzuo1976/mingpai-ai/raw/v2.1.0/temporal_agent_plugin/
v2.0.0
Added task pause/resume functionality, Bayesian timeout predictor for cold-start scenarios, and improved time drift detection
v1.0.0
智能体时序感知插件 1.0.0 — 初始版本发布
- 提供任务执行过程中的耗时预估、进度感知、异常检测等功能，缓解AI“黑盒焦虑”。
- 支持用户交互中的打断、沉默和上下文时效性管理，提升AI的社交时序感知。
- 实现物理/数字操作的动作时序对齐、长流程任务记忆和因果关系预测。
- 核心组件包含 TemporalAnalyzer、ProgressEstimator、AnomalyDetector、SocialTemporal 和 CausalPredictor。
- 集成适配 LangChain、LlamaIndex 及自定义AI框架。
- 开箱即用，适用AI助手、自动化工具、物理操作智能体等多种场景。
元数据
常见问题
temporal-agent-plugin 是什么？
Provides temporal awareness for AI agents, including timing prediction, progress monitoring, and social timing understanding. Invoke when AI needs time perce... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 131 次。
如何安装 temporal-agent-plugin？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install temporal-agent-plugin」即可一键安装，无需额外配置。
temporal-agent-plugin 是免费的吗？
是的，temporal-agent-plugin 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
temporal-agent-plugin 支持哪些平台？
temporal-agent-plugin 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 temporal-agent-plugin？
由 xmzuo1976（@xmzuo1976）开发并维护，当前版本 v2.1.0。
推荐 Skills