Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mrpulor-gh/nuphus --skill mcp-toolsgit clone --depth 1 https://github.com/mrpulor-gh/nuphusWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mrpulor-gh/nuphus/mcp-tools)<a href="https://agentmods.dev/skills/mrpulor-gh/nuphus/mcp-tools"><img src="https://agentmods.dev/badge/skills/mrpulor-gh/nuphus/mcp-tools/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mrpulor-gh/nuphus/mcp-tools"><img src="https://agentmods.dev/badge/skills/mrpulor-gh/nuphus/mcp-tools.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.00976 |
| Opus 5.5 | $0.00000 | $0.00390 |
| Sonnet 5.5 | $0.00000 | $0.00195 |
| Haiku 4.5 | $0.00000 | $0.00098 |
Grade A, and why
mcp-tools scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP 工具参考
通过
skill_read mcp-tools加载本文档。WorkAgent 设计工作流时按需查阅。
原理
Nuphus 通过 MCP (Model Context Protocol) 连接外部工具。每个 MCP server 暴露一组 tools，可在工作流中用 do.mcp 步骤调用（V2 格式，与内置工具同级别）：
{ "id": "search", "name": "搜问题",
"do": { "mcp": { "server": "github", "tool": "search_issues",
"with": { "query": "repo:nuphus bug", "limit": 5 } } },
"capture": "issues" }
server：plugin/mcp/servers.yaml中的 keytool：MCP 工具名with：工具参数（支持{{var}}模板）capture：字符串，输出存入变量（V2 规范，无as_type对象格式）
工具发现
servers.yaml 配置了新 server 但不确定其工具签名时，运行时自省：
{ "id": "list", "name": "列工具",
"do": { "mcp": { "server": "github", "tool": "tools/list", "with": {} } },
"capture": "tools" }
{{tools}} 含完整工具名与 inputSchema。
已配置的 Server
以下列表来自实际合并后的配置（plugin/mcp/servers.yaml + 本机私有覆盖）。配置变更后需同步更新本文档。
两层配置（合并读取，同名时本机层覆盖通用层）：
| 层 | 路径 | 用途 | 是否入库 |
|---|---|---|---|
| 通用层 | plugin/mcp/servers.yaml |
随仓基础配置，可随发布版携带 | 是 |
| 本机层 | <数据目录>/mcp/servers.local.yaml |
仅本机可用的 server（本机路径/私有端点/个人工具） | 否 |
<数据目录> 即 crate::utils::nuphus_data_dir()：Windows 为 %APPDATA%\.nuphus，
macOS 为 ~/Library/Application Support/.nuphus，Linux 为 ~/.local/share/.nuphus。
本机私有配置格式（与通用层完全一致）：
servers:
<server-key>:
command: "npx"
args: ["-y", "<package>", "mcp"]
timeout_ms: 120000
auto_start: false
当前本机已配置的 server 以运行时实际合并结果为准（list_mcp_servers 可查）。
常用 Server 速查
| Server | 环境变量 | 常用工具 |
|---|---|---|
GitHub @modelcontextprotocol/server-github |
GITHUB_PERSONAL_ACCESS_TOKEN |
search_repositories / search_issues / create_issue / get_file_contents |
Postgres @modelcontextprotocol/server-postgres |
DATABASE_URL |
query（执行 SQL） |
Slack @modelcontextprotocol/server-slack |
SLACK_BOT_TOKEN |
send_message / list_channels / get_channel_history |
Filesystem @modelcontextprotocol/server-filesystem |
参数 --directory <允许目录> |
read_file / write_file / list_directory / search_files |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago Changed · +21 lines 5c217319a635
- 24d ago Changed · -2 lines 7c117ca25f32
- 29d ago Changed · +2 lines 6f06b5d62501
- 1mo ago First seen · 61 lines · 0 tokens per session scan A 12ebd720846d
mcp-tools is a skill published in the GitHub repository mrpulor-gh/nuphus (89 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 976 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
desktop-gui-inspect
Full-desktop automation for targets that have no browser and no API at all — a legacy Java-based NMS client, a vendor's Windows-only configuration utility, a terminal emulator with no scriptable interface. Drives OpenClaw's ClawHub computer-use skill (Xvfb+XFCE virtual desktop, xdotool input automation) to read…
blender-agent
Drive Blender with the official Blender MCP for everything inside the app, and computer-use only for what has no API. Never read Blender through the accessibility tree.
computer-use
Desktop automation with computer-use-mcp. Use when controlling native macOS/Windows apps via accessibility, OCR, screenshots or scripting — not for pure git/file/http tasks.
analytics-agent
Answer a business question from an organisation's real dashboards — citing measured numbers, using the company's own metric definitions, and showing the person what you looked at.
computer-use-forms
Fill and operate form fields. Accessibility where the app exposes it, the page's own DOM in a browser, pixels only when neither is available.
computer-use-recovery
Recover from FocusFailure, useractive and wrong-window input when using computer-use-mcp.