AgentScope Skills
Skills for AgentScope and QwenPaw: framework API guidance, agent memory, schedules, browser control, mail and multi-agent flows.
- 135 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- Some skills teach the agent to run schedule commands and drive a browser.
- The mail skills give the agent access to messages through `himalaya`.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/agentscope-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Some skills teach the agent to run schedule commands and drive a browser.
- The mail skills give the agent access to messages through `himalaya`.
This is third-party code. Review the repository files before installing.
What it does
The AgentScope team repository collects skills for building on this framework and for the QwenPaw app. The `agentscope-skill` skill covers the AgentScope 2.x API, agent tools, multi-agent orchestration and service deployment, with reference notes and signature-checking scripts in `references` and `scripts`. The `nano-memory` skill teaches an agent to keep memory in files: daily logs and a long-term `MEMORY.md`. The remaining skills, built into QwenPaw, come in `-en` and `-zh` pairs: `cron` schedule management, attaching to a running Chrome over CDP, file reading, mail through `himalaya`, requesting plans from a stronger agent and the DingTalk channel.
Who it is for. Developers on AgentScope and QwenPaw users whose Claude Code or Cursor writes framework code and must know its current API.
Good fit when
- An agent builds an AgentScope 2.x app and must not invent API calls.
- An agent needs persistent memory between sessions kept in plain files.
- A multi-agent system on AgentScope is designed from scratch.
- A QwenPaw workflow needs schedules, mail or browser attachment.
Not a fit when
- You do not use AgentScope or QwenPaw.
- You want one standalone skill: several rely on `qwenpaw` commands and are useless outside the app.
- The agent must not get access to mail or the browser.
Example request
Use agentscope-skill: build a ReAct agent with two tools and show how to deploy it as a service.Limitations
The README lists only `agentscope-skill` and `nano-memory`, while the `skills` folder holds dozens of QwenPaw skills. Each folder must be copied separately: the README `cp -r` command handles one skill. Several skills depend on QwenPaw app commands and on external programs such as `himalaya`. The mail and browser skills expose personal data to the agent, so enable them deliberately.
How to disable. Delete the skill folder from `~/.claude/skills/` or the project `.claude/skills`. `nano-memory` memory stays in `MEMORY.md` and `memory/` files, which you clean by hand.
FAQ
How do I install an AgentScope skill in Claude Code?
Clone the repository and copy the folder with `cp -r agentscope-skills/skills/{skill_name} ~/.claude/skills/`. For a project use `.claude/skills` instead of the home folder.
Does it work in Cursor?
Yes. Copy the needed skill into `.cursor/skills`, as with Claude Code.
How do the `-en` and `-zh` versions differ?
In the language of the instructions. The rest of each pair is the same, so take the one your agent needs.
What does nano-memory do?
It gives the agent a memory layout: a `memory/YYYY-MM-DD.md` log and a distilled `MEMORY.md`, kept with plain file commands.
Which skills exist beyond the two in the table?
The `skills` folder holds `cron`, `browser_cdp`, `browser_visible`, `file_reader`, `himalaya`, `make_plan`, `multi_agent_collaboration`, `news`, `dingtalk_channel` and more, each in `-en` and `-zh` versions.
Discussion
No questions or reviews yet.
Related
SkillsSkill for Claude Code and Cursor that teaches reactive AgentScope Java code and steers clear of common dependency mistakes.
A catalog of nearly two hundred skills: code delivery, review, debugging, AI governance, product work and agent coordination.
A collection of 14 portable skills for long agent runs: work planning, live app QA, search across past sessions and a phone call
A command-line skill manager: installs skills from a marketplace, GitHub, private Git and npm into 45 agents' folders and scans them first
A base AGENTS.md and 12 skills: delegation to Codex, Grok, Devin and Antigravity, planning, copywriting, scoped changes.
A personal collection of about 130 Claude Code and Codex skills: media pipelines, Cloudflare, YouTube, design and task review
README in short
The README describes the repository as a set of self-contained knowledge packs about the AgentScope ecosystem and the QwenPaw app, ready to install in Claude Code and Cursor. The table names two skills; the rest were added to `skills` later.
README badge
Original SKILL.md
--- name: agentscope-skill description: Build and debug Python applications using AgentScope 2.x. Consult this skill for AgentScope APIs, agent tools, multi-agent orchestration, and agent service deployment. metadata: version: "0.2.0" --- # AgentScope 2.0 AgentScope is an open-source framework for building and serving LLM-powered agent applications, from a single tool-using agent to coordinated multi-agent systems. It provides application orchestration and service infrastructure; model inference comes from configured providers, and tools execute through configured local or sandbox backends. It consists of two layers: - **Agent SDK:** Building blocks for agent applications, including agents, models, messages, tools, context and state management, middleware, memory, RAG, multi-agent orchestration, and workspaces. - **Service:** A service layer built on the SDK, providing APIs for agent and session management, persistence, teams, scheduling, channels, and resource management, with a Web UI example. This skill supports **AgentScope 2.x** and is based on **2.0.8**. API signatures and behavior should follow the SDK version actually installed in the user's environment. `agentscope-runtime` and `agentscope-studio` are not compatible with 2.x; use the built-in service and workspace capabilities instead. ## Installation Python **3.11 or newer** is required. ```bash pip install agentscope # or uv pip install agentscope ``` ## Core Concepts and Basic Example - `Agent` owns the reasoning/acting loop. Use `reply()` for a final `Msg`, or `reply_stream()` for events. `launch_console()` handles terminal interaction, tool confirmation, and interruption. - Construct provider models with a credential object and `model=...`. Formatters still exist, but are configured on the model; providers select a default formatter. They are not passed to `Agent`.