FAQ & Community

Frequently asked questions, roadmap, community links, and how to contribute to CoPaw.

FAQ

This page collects the most frequently asked questions from the community. Click a question to expand the answer.

CoPaw vs OpenClaw: Feature Comparison

Please check the Comparison page for detailed feature comparison.

How to install CoPaw

CoPaw supports multiple installation methods. See Quick Start for details:

  1. One-line installer (sets up Python automatically)
# macOS / Linux:
curl -fsSL https://copaw.agentscope.io/install.sh | bash
# Windows (PowerShell):
irm https://copaw.agentscope.io/install.ps1 | iex
# For latest instructions, refer to docs and prefer pip if needed.
  1. Install with pip

Python version requirement: >= 3.10, < 3.14

pip install copaw
  1. Install with Docker

If Docker is installed, run the following commands and then open http://127.0.0.1:8088/ in your browser:

docker pull agentscope/copaw:latest
docker run -p 8088:8088 -v copaw-data:/app/working agentscope/copaw:latest

How to update CoPaw

To update CoPaw, use the method matching your installation type:

  1. If installed via one-line script, re-run the installer to upgrade.
  2. If installed via pip, run:
pip install --upgrade copaw
  1. If installed from source, pull the latest code and reinstall:
cd CoPaw
git pull origin main
pip install -e .
  1. If using Docker, pull the latest image and restart the container:
docker pull agentscope/copaw:latest
docker run -p 8088:8088 -v copaw-data:/app/working agentscope/copaw:latest

After upgrading, restart the service with copaw app.

How to initialize and start CoPaw service

Recommended quick initialization:

copaw init --defaults

Start service:

copaw app

The default Console URL is http://127.0.0.1:8088/. After quick init, you can open Console and customize settings. See Quick Start.

Open-source repository

CoPaw is open source. Official repository: https://github.com/agentscope-ai/CoPaw

Where to check latest version upgrade details

You can check version changes in CoPaw GitHub Releases.

How to configure models

In Console, go to Settings -> Models. See Console -> Models for details.

  • Cloud models: fill provider API key (ModelScope, DashScope, or custom), then choose the active model.
  • Local models: supports llama.cpp, MLX, and Ollama. After download, select the active model on the same page.

You can also use copaw models CLI commands for configuration, download, and switching. See CLI -> Models and Environment Variables -> copaw models.

How to manage Skills

Go to Agent -> Skills in Console. You can enable/disable Skills, create custom Skills, and import Skills from Skills Hub. See Skills.

How to configure MCP

Go to Agent -> MCP in Console. You can enable/disable/delete/create MCP clients there. See MCP.

Common error

1. Error pattern: You didn't provide an API key

Error detail:

Error: Unknown agent error: AuthenticationError: Error code: 401 - {'error': {'message': "You didn't provide an API key. You need to provide your API key in an Authorization header using Bearer auth (i.e. Authorization: Bearer YOUR_KEY). ", 'type': 'invalid_request_error', 'param': None, 'code': None}, 'request_id': 'xxx'}

Cause 1: model API key is not configured. Get an API key and configure it in Console -> Settings -> Models.

Cause 2: key is configured but still fails. In most cases, one of the configuration fields is incorrect (for example base_url, api key, or model name).

CoPaw supports API keys obtained via DashScope Coding Plan. If it still fails, please check:

  • whether base_url is correct;
  • whether the API key is copied completely (no extra spaces);
  • whether the model name exactly matches the provider value (case-sensitive).

Reference for the correct key acquisition flow: https://help.aliyun.com/zh/model-studio/coding-plan-quickstart#2531c37fd64f9

How to get support when errors occur

To speed up troubleshooting and fixes, please create an issue in the CoPaw GitHub repository and include complete error information: https://github.com/agentscope-ai/CoPaw/issues

In many Console errors, a detailed error file path is included. For example:

Error: Unknown agent error: AuthenticationError: Error code: 401 - {'error': {'message': "You didn't provide an API key. You need to provide your API key in an Authorization header using Bearer auth (i.e. Authorization: Bearer YOUR_KEY). ", 'type': 'invalid_request_error', 'param': None, 'code': None}, 'request_id': 'xxx'}(Details: /var/folders/.../copaw_query_error_qzbx1mv1.json)

Please upload that file (for example /var/folders/.../copaw_query_error_qzbx1mv1.json) together with your current model provider, model name, and exact CoPaw version.

Roadmap

Roadmap table

Area Item Status
Horizontal Expansion More channels, models, skills, MCPs — community contributions welcome Seeking Contributors
Existing Feature Extension Display optimization, download hints, Windows path compatibility, etc. — community contributions welcome Seeking Contributors
Compatibility & Ease of Use App-level packaging (DMG, EXE) In Progress
One-click deployment: built-in deps, dev extras, install/upgrade tutorials In Progress
Release & Contributing Contributing docs and test framework In Progress
Responsive handling of community contributions In Progress
Contributing guidance for vibe coding agents Planned
Bugfixes & Enhancements Message collapse/hide in UI Planned
Skills and MCP runtime install, hot-reload improvements Planned
Context management and compression (long tool outputs, lower token usage) Planned
Multimodal support In Progress
Security Shell execution confirmation Planned
Tool/skills security Planned
Configurable security levels (user-configurable) Planned
Multimodal Voice/video calls and real-time interaction Long-term planned
Multi-agent Built on AgentScope, native multi-agent workflows Long-term planned
Sandbox Deeper integration with AgentScope Runtime sandboxes Long-term planned
Self-healing Daemon agent for automated recovery and health monitoring Long-term planned
CoPaw-optimized local models LLMs tuned for CoPaw's native skills and common tasks; better local personal-assistant usability Long-term planned
Small + large model collaboration Local LLMs for sensitive data; cloud LLMs for planning and coding; balance of privacy, performance, and capability Long-term planned
Cloud-native Deeper integration with AgentScope Runtime; leverage cloud compute, storage, and tooling Long-term planned
Skills Hub Enrich the AgentScope Skills repository and improve discoverability of high-quality skills Long-term planned

Status: In Progress — actively being worked on; Planned — queued or under design, also welcome contributions; Seeking Contributors — we strongly encourage community contributions; Long-term planned — longer-horizon roadmap.

Get involved

We are building CoPaw in the open and welcome contributions of all kinds! Check the table above (especially items marked Seeking Contributors) to find areas that interest you, and read CONTRIBUTING to get started. We particularly welcome:

  • Horizontal expansion — new channels, model providers, skills, MCPs.
  • Existing feature extension — display and UX improvements, download hints, Windows path compatibility, and the like.

Join the conversation on GitHub Discussions to suggest or pick up work.

Bug reports & community

Report bugs, issues, or get in touch with the team through either channel below.

Join the community

Discord DingTalk
Discord DingTalk

Open Source & Contribution

CoPaw is open source. The project repository is hosted on GitHub:

https://github.com/agentscope-ai/CoPaw

How to Contribute

We welcome contributions including bug reports, feature ideas, documentation improvements, and code changes. You can open issues and pull requests on the repository above. Contribution guidelines, code of conduct, and participation workflow are available in the repo — we look forward to your participation!

We also encourage you to contribute according to our Roadmap: items marked Seeking Contributors (e.g. new channels, model providers, skills, MCPs, or display/UX improvements) are great places to start. Check the roadmap for current focus and long-term plans, then join the conversation on GitHub Discussions.

Previous ← Feature Comparison