← 返回 Skills 市场

日常小助手 MCP Server

Name: 日常小助手 MCP Server
Author: asuranale

作者 AsuraNale · GitHub ↗ · v2.1.0 · MIT-0

cross-platform ✓ 安全检测通过

180

总下载

当前安装

版本数

在 OpenClaw 中安装

/install daily-assistant

功能描述

AI-powered daily task management MCP Server — recommend next task, inherit uncompleted todos, detect overdue, generate reviews. Deterministic ops in code (ze...

使用说明 (SKILL.md)

\r \r

Daily Assistant MCP Server\r

\r A personal task management MCP Server. Deterministic operations (parsing, sorting, statistics) run in Python with zero token cost — AI only steps in when creativity is needed.\r \r

Setup\r

git clone https://github.com/AsuraNale/daily-assistant-mcp.git\r
cd daily-assistant-mcp\r
python3 src/setup.py --auto   # Windows: py src/setup.py --auto\r
```\r
\r
The setup wizard creates a `.venv`, installs dependencies, sets up your data directory, and auto-configures your AI editor. No manual `pip install` or config editing needed.\r
\r
## Tools\r
\r
| Tool | What it does |\r
|------|-------------|\r
| `recommend_next` | Recommends the most important task to work on next |\r
| `get_today` | Reads today's daily task file |\r
| `inherit_tasks` | Carries over uncompleted tasks from yesterday |\r
| `check_overdue` | Detects overdue task files |\r
| `generate_review` | Generates end-of-day review with completion stats |\r
| `scan_split` | Flags tasks >80min or missing time estimates |\r
\r
## Resources\r
\r
| Resource | Content |\r
|----------|---------|\r
| `daily://today` | Today's task file |\r
| `daily://dashboard` | Dashboard overview |\r
| `daily://history` | 7-day completion statistics |\r
\r
## Task Format\r
\r
```markdown\r
- [ ] Task description ⏱️45min 📅 2026-03-30 ⏫\r
```\r
\r
Markers: `⏱️` = time estimate, `📅` = deadline, `⏫` = highest priority, `🔼` = high, `🔽` = low.\r
\r
## Design Philosophy\r
\r
- **Zero-token deterministic ops**: Parsing, sorting, stats run in Python code\r
- **AI only when needed**: Task splitting, creative advice, context-aware suggestions\r
- **Platform-independent**: Windows, macOS, Linux. No Obsidian dependency\r
- **Simple data**: Plain Markdown files, edit anywhere\r

安全使用建议

What to consider before installing: - This repo will create a .venv and run pip install fastmcp (network access to PyPI). If you prefer, inspect setup.py and run the venv/pip steps manually. - setup.py will create a data folder (default: ~/Desktop/日常小助手/Daily) and write Dashboard.md and config.json. If you want a different location, run setup interactively instead of --auto. - setup.py auto-merges an entry into various editor MCP config files (e.g., ~/.claude.json, ~/.cursor/mcp.json, ~/.kiro/..., ~/.codeium/...). Those files may already contain API keys or other settings; the script attempts a safe merge but you should back up those config files before running --auto. - The SKILL.md suggests git cloning the repo from GitHub; the registry already contains the code — cloning fetches the upstream repo (possible update). If you want zero-network install, skip the git clone and run setup.py from the bundled files after reviewing them. - I reviewed the visible parts of core.py, server.py and setup.py; the logic is consistent with the stated purpose (file parsing, ranking, writing daily files, and adding MCP config). core.py was truncated in the provided bundle, so if you need higher assurance, inspect the full core.py for any unexpected network calls or obfuscated behavior. - Recommended actions: (1) review setup.py and server.py locally before running, (2) back up editor config files you care about, (3) run setup in interactive mode (no --auto) to see prompts and paths, or run inside an isolated account/VM if you want stronger containment.

功能分析

Type: OpenClaw Skill Name: daily-assistant Version: 2.1.0 The daily-assistant skill bundle is a legitimate task management MCP server. It provides tools for parsing, ranking, and reviewing daily tasks stored in Markdown files. The setup script (setup.py) automates the creation of a virtual environment and the configuration of AI editors (Claude Code, Cursor, Kiro, Windsurf) by safely merging MCP server settings into their respective local configuration files. The core logic (core.py) and server implementation (server.py) use standard Python libraries and follow safe coding practices, with no evidence of data exfiltration, malicious execution, or prompt injection.

能力评估

✓ Purpose & Capability

The name/description (daily task MCP server) matches the actual code and instructions: Python-based deterministic task parsing, tools to read/write daily Markdown files, recommendations, overdue checks, and integration with MCP-compatible editors. Requiring python3 and creating a venv + installing fastmcp is proportional to this purpose. The only notable capability beyond bare file ops is that the setup will modify editor configuration files (to register the MCP server) — this is explainable by the goal of integrating with editors.

ℹ Instruction Scope

SKILL.md tells the user to git clone and run src/setup.py --auto. setup.py will create a .venv, pip-install fastmcp, create a data directory (~/Desktop/日常小助手/Daily), generate config.json, and auto-merge an entry into various editor MCP config files (e.g. ~/.claude.json, ~/.cursor/mcp.json, ~.kiro/settings/mcp.json, etc.). These actions go beyond merely reading/writing task files because they modify editor config files in the user's home directory; this is consistent with registering an MCP server but worth explicit user consent.

ℹ Install Mechanism

No registry install spec is provided, but the package's setup.py will create a virtualenv and run pip install fastmcp (download from PyPI). The SKILL.md also instructs a git clone from the project's GitHub. Both git clone (GitHub) and pip install from PyPI are typical but require network access and trust in upstream sources. There is no download from obscure URLs or archive extraction from untrusted hosts.

ℹ Credentials

The skill does not request environment variables or credentials. However, setup.py reads and writes editor config files under the user's home (which may contain API keys/tokens for those editors). The script attempts a safe JSON merge (preserving existing content) but writing into files that can contain secrets is a sensitive action — the change is explainable by the integration goal but should be done with caution and backups.

ℹ Persistence & Privilege

always:false (no forced global installation). The setup will register the server persistently in detected editor MCP config files, making the server available to editors in the future. That persistent registration is a normal part of integrating an MCP server, but it does grant ongoing presence in editor configs and should be done only with user consent. The skill does not request elevated OS privileges.

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 daily-assistant
安装完成后，直接呼叫该 Skill 的名称或使用 /daily-assistant 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

v2.2.0

Web interface added for daily-assistant with new front-end assets. - Introduced web server (src/web.py) for browser-based interaction. - Added static files: app.js, index.html, and style.css for the front end. - Updated README and SKILL.md to reflect web interface availability. - Existing Python core and setup scripts updated to support the new web functionality.

v2.1.0

- Added new features and improvements across core, server, and setup modules. - Updated version to 2.1.0. - Documentation refreshed for clarity, with enhanced setup and usage instructions. - Improved install and configuration automation via setup script.

v2.0.0

Task deduplication: inherited tasks marked as [>] are excluded from overdue/history queries

v1.0.1

Remove test files and internal project documents from published package

v1.0.0

Initial release of Daily Assistant MCP Server. - AI-powered daily task manager with deterministic Python operations (no token cost) - Recommends next task, inherits unfinished tasks, checks for overdue, generates reviews - Simple setup for Windows/macOS/Linux; no external dependencies like Obsidian - Tasks and dashboards use easy-to-edit plain Markdown files - Provides tools for daily reviews, detecting overdue tasks, and intelligent task recommendations

元数据

Slug daily-assistant

版本 2.1.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 5

常见问题