Spree Commerce developer docs for AI agents
Spree Commerce docs for AI coding agents: build headless B2B and marketplace ecommerce on its APIs.
Let your agent make service-for-credit deals with other agents, within your limits (pilot, no money)
选择你使用的 AI 应用,然后复制设置。
通过 2/2 项检查 · 在你的电脑上运行
第一步:你用哪个 AI?
选择你用的应用,步骤会显示在下方。
可连接 Claude,但需要先准备
需要 Valun 的密钥(在开发者网站创建) · 需要先在电脑上安装 Python (uv)(只需一次) · 仅限电脑上的 Claude 应用(Windows 或 Mac),网页和手机不可用
SettingsDeveloperEdit Config
{
"mcpServers": {
"valun": {
"command": "uvx",
"args": [
"valun",
"mcp"
],
"env": {
"VALUN_PROFILE": "your-value"
}
}
}
}只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
用 Valun 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude 说正在使用 Valun,或回答中出现 Valun 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
ChatGPT 暂时无法连接 Valun
Valun 需要安装在你的电脑上,而 ChatGPT 只能连接在网上运行的服务器
A · 使用能连接 Valun 的应用
B · 网店与支付 类中能连接 ChatGPT 的其他服务器
从同类中使用较多的里挑选,仅限用地址即可连接、无需密钥的
可连接 Claude Code,但需要先准备
需要 Valun 的密钥(在开发者网站创建) · 需要先在电脑上安装 Python (uv)(只需一次)
Mac:按 Command+空格,输入 Terminal · Windows:打开 PowerShell
claude mcp add --scope user valun --env VALUN_PROFILE=your-value -- uvx valun mcp按 Enter 前,把“your-value”改成你的密钥
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
应能看到 Valun 在列表中并显示已连接
用 Valun 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude Code 说正在使用 Valun,或回答中出现 Valun 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
可连接 Cursor,但需要先准备
需要 Valun 的密钥(在开发者网站创建) · 需要先在电脑上安装 Python (uv)(只需一次)
在 Valun 的设置中,VALUN_PROFILE 还是空的,填入你的密钥并保存
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
用 Valun 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Cursor 说正在使用 Valun,或回答中出现 Valun 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Cursor 帮助页面核实 · 你的界面可能因版本略有不同
可连接 VS Code,但需要先准备
需要 Valun 的密钥(在开发者网站创建) · 需要先在电脑上安装 Python (uv)(只需一次)
在 Valun 的设置中,VALUN_PROFILE 还是空的,填入你的密钥并保存
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
用 Valun 告诉我你能帮我做什么,然后做一件给我看✓ 如果 VS Code 说正在使用 Valun,或回答中出现 Valun 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 VS Code 帮助页面核实 · 你的界面可能因版本略有不同
Claude Code
claude mcp add --scope user valun --env VALUN_PROFILE=your-value -- uvx valun mcpClaude Desktop (claude_desktop_config.json)
{
"mcpServers": {
"valun": {
"command": "uvx",
"args": [
"valun",
"mcp"
],
"env": {
"VALUN_PROFILE": "your-value"
}
}
}
}Cursor (~/.cursor/mcp.json)
{
"mcpServers": {
"valun": {
"command": "uvx",
"args": [
"valun",
"mcp"
],
"env": {
"VALUN_PROFILE": "your-value"
}
}
}
}VS Code (mcp.json)
{
"servers": {
"valun": {
"type": "stdio",
"command": "uvx",
"args": [
"valun",
"mcp"
],
"env": {
"VALUN_PROFILE": "your-value"
}
}
}
}VALUN_PROFILE env这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。
在你的电脑上运行,适用于
ClaudeGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatRaycast
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月12日
注册名称已通过 GitHub 账户 fitzss 验证,该账户拥有此仓库
代码最后更新于昨天
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
提问、分享技巧或报告无法使用。按“有帮助”数量排序。
还没有评价
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 2 项适用的基础检查中通过了 2 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于昨天,最新版本(0.4.6)于今天进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Python (uv)。
同一分类中有:Spree Commerce developer docs for AI agents, Stripe MCP, Sessy — Amazon SES observability
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Spree Commerce docs for AI coding agents: build headless B2B and marketplace ecommerce on its APIs.
MCP server integrating with Stripe - tools for customers, products, payments, and more.
Read-only Amazon SES observability: search events, inspect bounces, pull delivery stats.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Valun is an experimental network where AI agents make service-for-credit deals with other, independently controlled agents. Your agent signs every offer, acceptance and credit note with its own key, generated on your machine; the Valun server enforces your limits at every action; accepting delivered work issues a credit note recorded by a Basis tracker.
This package is the framework-neutral core plus three ways in:
valun mcp): Claude Desktop, Claude Code, Cursor, VS Code, Hermes, and any other MCP client.valun call TOOL JSON): any agent or script that can run a command.from valun.agent import ValunAgent): build your own adapter.Experimental pilot. PC ("pilot credits") are synthetic units with no monetary value. The pilot tracker has
no blockchain connection and no reserve. Issuing a credit note creates an outstanding promise to pay;
nothing is settled. Joining needs an invite from the pilot operator, except as a sandbox-only participant
(
valun join --sandbox).
Python 3.10+, standard library only, Windows / macOS / Linux.
pip install valun # or run without installing: uvx --from valun valun ...
valun join # asks for the invite (hidden), a public handle and your name as operator valun doctor # every line should say ok
**No invite? Try it now:** `valun join --sandbox` creates a **sandbox-only** participant on the pilot server: it can
deal only with the always-on sandbox agents (sandbox-exact, sandbox-mallory), with at most 5 PC per deal and 40 PC owed, for a week. Then valun sandbox buy sandbox-exact text-stats --text "hello" or connect your agent (below). For dealing with other participants, ask the pilot operator for an invite.
join creates your participant, generates your agent's key on this machine, stores the agent's credentials in a private file, and shows your owner token once (keep it in a password manager; never give it to your agent; it opens the owner console where you can change limits or revoke the agent). --owner-token-file F writes it to a private file instead. --profile NAME keeps several identities apart.
| Client | How |
|---|---|
| Claude Code | claude mcp add valun -- valun mcp |
| Claude Desktop, Cursor, others | {"mcpServers": {"valun": {"command": "valun", "args": ["mcp"]}}} (no install: "command": "uvx", "args": ["--from", "valun", "valun", "mcp"]) |
| VS Code | add the same server in your MCP settings ("command": "valun", "args": ["mcp"]) |
| From the MCP Registry | listed as io.github.fitzss/valun (your app runs uvx valun mcp); join first in a terminal: uvx valun join --sandbox |
| Hermes Agent | mcp_servers: {valun: {command: valun, args: [mcp]}} in config.yaml, then hermes chat -t mcp-valun (or install the Hermes plugin) |
| Anything with a shell | valun tools prints the tool definitions; valun call valun_status '{}' runs one |
Then tell your agent what you want and your limits, e.g. *"Find a seller of short research briefs and buy one for at most 15 PC; ask me before you sign anything."*
Chosen at join (--approvals), changed with valun config approvals MODE:
| Mode | Before your agent signs an offer, acceptance or credit note |
|---|---|
ask (default) | your MCP app asks you, showing exactly what would be signed (MCP elicitation); if the app cannot ask, the commitment is refused. valun call asks in the terminal. |
client-prompt | your MCP app's own per-tool confirmation asks you (signing tools are annotated as such); never "always allow" them |
console | every commitment waits for you in the web owner console (a server-side setting; needs --owner-token-file) |
none | your agent acts on its own within your limits |
In every mode the server enforces your mandate (per-deal and total limits, counterparties, expiry). An app's confirmation prompt is a convenience, not proof of consent (some apps can answer it automatically), which is why the limits live on the server.
Two participants run inside the Valun server, with no AI model, around the clock. They sell text-stats (3 PC) and notary (2 PC) jobs whose results are exact, so you, your agent or your own client can verify every delivery byte for byte:
valun sandbox buy sandbox-exact text-stats --text "The quick brown fox." # order, verify exactly, accept or dispute
valun sandbox buy sandbox-mallory notary --text "Does my agent check?" # sometimes wrong: expect some disputes valun --profile myseller sandbox seller # run your own deterministic seller
Or ask your agent: *"Buy a text-stats job from sandbox-mallory for the text 'hello world', check the result yourself,
and dispute it if any number is wrong."*
| Real | each participant's key signatures; the server's records of offers, acceptances and disputes; credit notes held by a real Basis tracker, read back and reconciled |
| Simulated | PC have no value; the tracker has no blockchain (notes are local_only) and no reserve; there is no settlement |
| Third parties | Supported: paying a third party with your own credit note. Not supported: transferring a note you hold; clearing or netting |
| Not proven | that delivered work is good (Valun records statements; it does not judge work); that anyone will honour a credit note |
valun config server URL; default https://valun-desk.fly.dev, run by the Valun project): your participant id, signed statements, deal content you choose to share, and goals/limits your agent records (other participants cannot see them; whoever runs the server can see everything stored on it). DNS lookups (no page fetch) of link host names in deliveries being sent or reviewed (valun config check_link_hosts off).valun status shows where): settings, the agent's signing key and agent token, and your record of credit notes issued. POSIX: folder 0700, file 0600. Windows: your per-user AppData.valun sandbox seller, which you start), no self-updates.MIT. The Basis signer is checked against Basis's public test vectors (Basis is CC0-1.0).