OSF CLI Go
Read-only MCP tools for authenticated Open Science Framework projects, components, files, and contributors.
README
osf-cli-go
A Go command-line client for the Open Science Framework (OSF).
Features
osf auth whoami— Identify the authenticated OSF accountosf auth login— Guided personal-access-token bootstrap for username/password usersosf projects list|get— List and inspect projectsosf components list— List project componentsosf files list|download|upload|mkdir|rm— Browse, download, upload, create folders, and delete OSF Storage filesosf searchandosf preprints list— Search OSF and list preprintsosf registrations create— Create draft registrations for an existing nodeosf export— Export a node snapshot as JSON or a summary tableosf-mcp— Stdio MCP server exposing read-only OSF tools for agent clientsosf completion bash|zsh|fish|powershell— Shell completion scripts- JSON and human-readable output modes
- Safe, atomic file downloads with conflict policy (fail/skip/overwrite)
Install
Requirements:
- Go 1.26 or newer
go install github.com/edithatogo/osf-cli-go/cmd/osf@latest
go install github.com/edithatogo/osf-cli-go/cmd/osf-mcp@latest
Or from a local checkout:
go build -o bin\osf.exe ./cmd/osf
go build -o bin\osf-mcp.exe ./cmd/osf-mcp
.\scripts\build.ps1
Authentication
Set OSF_TOKEN in your shell session. Do not commit the token or write it into project files. OSF_USERNAME and OSF_PASSWORD are supported as an opt-in fallback credential source, but personal access tokens remain preferred for automation and for accounts using SSO or two-factor authentication.
$env:OSF_TOKEN = '<your-token>'
For guided token setup:
osf auth login
Quick Start
osf --help
osf auth whoami
osf projects list
osf projects get https://osf.io/abc12/
osf components list abc12
osf files list abc12
osf files download --file <file-id> ./output/
osf files download --tree abc12 ./output/
osf files upload --node abc12 ./report.pdf
osf search "open science"
osf preprints list
osf registrations create abc12 --schema <schema-id> --title "Analysis plan"
osf export abc12 --json
MCP Server
osf-mcp runs a stdio MCP server with read-only tools:
osf_whoami, osf_projects_list, osf_project_get,
osf_components_list, osf_files_list, and osf_contributors_list.
Local development configs are included for GitHub Copilot, VS Code, Claude,
Codex, Gemini CLI, and Qwen Code. Public registry metadata is in server.json
and registry/.
Related MCP Servers
Other maintained servers in the same Glama portfolio:
- FYI MCP for freedom-of-information request workflows.
- Healthpoint MCP for licensed health-service directory data.
- SourceRight for reference and citation verification.
Related OSF ecosystem servers:
- OSF connector by pipeworx-io for hosted OSF connectivity.
- Paperclip for multi-provider scholarly and OSF Preprints search.
Output Modes
All commands support --output table|json and --json shorthand:
osf projects list --json
osf auth whoami --output json
Project Status
The CLI is offline-tested for read-only operations, file downloads, WaterButler write primitives, search, preprint listing, draft registration creation, project create/update/delete operations, and node export. All Conductor tracks are reconciled against their per-track plans with closeout review evidence; live OSF validation remains opt-in because it requires credentials and network access.
Documentation
- Release checklist
- Documentation site source
- Install guide
- Usage guide
- Command reference
- Examples
- Architecture
- Developer guide
- MCP roadmap
- Contributing
License
Apache 2.0 — see LICENSE.
Citation
If you use this software in your research, please cite it using the metadata in CITATION.cff.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。