| name | clawbox |
| version | 1.0.0 |
| description | Use the ClawBox CLI to upload, download, organize, search, and share files on ClawBox (clawbox.ink) or a self-hosted ClawBox server. Trigger this skill when users mention clawbox, clawbox, cloud file storage, uploading files to the cloud, semantic file search, file sharing links, organizing files into folders, or managing tokens and storage quotas. Also use when code imports or references the clawbox CLI tool. |
ClawBox
ClawBox — a cloud file system for agents with token-based auth, semantic search, folders, and file sharing.
Setup
If the CLI is not installed, run the setup script:
bash ClawBoxSkill/scripts/setup.sh
Or manually:
git clone https://github.com/Alfra-AI/clawbox.git
cd clawbox
pip install .
clawbox init
Commands
clawbox init
clawbox init --api-url http://localhost:8000
clawbox config --show
clawbox upload <file>
clawbox upload <file> --path /docs/file.pdf
clawbox download <file_id>
clawbox list
clawbox list --folder /docs/ --recursive
clawbox search "query"
clawbox embed <file_id>
clawbox embed --failed
clawbox delete <file_id>
clawbox status
Workflow
- Check if
clawbox is installed. If not, run the setup script.
- Run
clawbox status to verify server connection and token.
- If no token, run
clawbox init.
- Prefer non-destructive inspection before mutation:
status, config --show, list.
- Ask before deletion unless the user explicitly requested it.
- Report key output after each operation: file id, download path, share link, etc.
Key details
- Config:
~/.clawbox/config.json
- Default server:
https://clawbox.ink
- Searchable formats: text, JSON, XML, PDF, Word, Excel, PowerPoint, CSV
- Storage: 10 MB free (anonymous), 1 GB with Google login
- Folders: virtual paths like
/docs/reports/, created implicitly on upload
For detailed API reference, read ClawBoxSkill/references/api.md.
For self-hosting and deployment, read ClawBoxSkill/references/self-hosting.md.
For troubleshooting, read ClawBoxSkill/references/troubleshooting.md.