list_files
List files in the allowed folder.
無需輸入參數。

io.github.techsaurabh-394由 GitHub 帳號 techsaurabh-394 發布v1.5.0更新於 Nov 16, 2026
File server whose delete tool asks a human first (elicitation or one-time token), with audit log.
發布方自己的服務:actions.world
MCP Safe Actions 的本機 MCP 服務;無需登入;STDIO;2 個工具。
暫無 AI 概覽:以下資訊來自註冊中心條目和服務本身。
Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。
參照 儲存庫 中的啟動說明。
[mcp-safe-actions: human approval for destructive MCP tools]
[CI] [101 tests passing] [100% line coverage] [TypeScript strict] [MIT]
[圖片] Star this repo if it saves you from a bad tool call · [Buy me a coffee]
Your AI agent should not be able to delete a file, send an email or drop a table just because it decided to.
mcp-safe-actions is a small TypeScript library for people who build MCP servers (Model Context Protocol). Wrap a dangerous tool once and it does nothing until a human approves. The model cannot approve its own action.
delete_file("../../important"). "Please ask first" in a prompt is a request, not a control.This library packages that pattern once, with a test suite that proves the safety rules, instead of rebuilding it in every server.
It also ships a small file server where deleting needs a human OK. Add it to your MCP client config:
Tools: list_files and delete_file (asks you first). Set AUDIT_LOG to keep a JSONL log. Listed in the official MCP Registry as io.github.techsaurabh-394/mcp-safe-actions.
Published on npm: https://www.npmjs.com/package/mcp-safe-actions
You also need @modelcontextprotocol/sdk (1.17+) and zod (you probably have both). Node 18+.
Replace server.registerTool(...) with registerConfirmedTool(server, ...) and split the work into preview (describe, don't change anything) and execute (runs after approval):
The tool is listed with destructiveHint: true and a description telling the model to wait for the user.
Claude Desktop (claude_desktop_config.json), using the included example server and a throwaway folder:
Claude Code: claude mcp add files -- node /absolute/path/to/mcp-safe-actions/dist/examples/files-server.js
Run npm run build first. Then ask the assistant to delete a file in the test folder. You should see either an approval prompt (clients with elicitation) or a preview with a token and a question from the assistant (clients without). The file must still exist until you say yes.
[Flow: model calls, server previews, human approves, action executes, audit logged]
Client supports elicitation (best case):
Client without elicitation (token fallback):
What the tests prove, among other things: a declined prompt never runs the action; a model-supplied token cannot skip the human prompt; tokens are single-use, expire, and only work for the same tool and arguments; two simultaneous uses of one token run the action exactly once; a prompt that times out or crashes fails closed; secret argument values never reach the audit log; the example server blocks path traversal and symlink escapes.
Details and an honest note about a weak test the mutation check caught: docs/TESTING.md.
Be clear-eyed:
mode: "elicit".execute() should re-check state that may have changed since.Full threat model: docs/SECURITY-MODEL.md.
For a screen-control tool, click(element_id) is harmless or a payment depending on the element. Give the tool a resolve function that looks the element up, and a gate for the labels you never want unattended:
The screen can change between approval and execution (a dialog opens, a list re-sorts), so resolve runs a second time right before execute. If the resolved context is different from the one the human approved, nothing runs, the audit log gets a context-changed event, and the model is told to ask again. A failed second lookup also refuses. In the token flow the token is bound to the resolved target as well. By default every returned field is compared; set pinFields: ["label", "role", "url"] to leave out volatile ones like coordinates.
Rules, all fail closed: a gate match always asks a human, even if autoApprove says yes. If resolve throws, rejects or returns anything other than a plain object (for example undefined for an unknown element), the call is never auto-approved. The approval prompt shows the resolved target (target label: Buy now) so the person sees what they are approving. Labels can be misleading (a button called "Continue" can submit a payment), so treat the gate as a first filter and also resolve the destination (url) for irreversible actions.
confirm_token is a reserved input name. More in docs/USAGE.md, including how to test your own tools with an in-memory client.
Add the same safety net to your own server's repo:
test/helpers.ts in this repo shows how to simulate a human who approves, declines, or never answers.
mode: "channel" + webhookApprover (point it at a service that can reach Slack or a dashboard)autoApprove lets low-risk calls run, everything else still asks a humanresolve looks up what a call targets (e.g. the label behind click(element_id)) and gate forces a human decision for labels like buy / delete / sendresolve runs again right before execute and the call is refused if the target changed after approvalIssues and PRs welcome: CONTRIBUTING.md. Any change to approval logic needs a test showing the action does not run without approval, and npm run mutation must still pass.
If this helped you ship a safer agent, please [圖片] star the repo, share it, or buy me a coffee:
MIT. See LICENSE. Copyright (c) 2026 Saurabh Pathak.
Saurabh Pathak · AI Engineer & Full-Stack Developer
GitHub ·
LinkedIn ·
Blog ·
[email protected]
來源:README.md,提交 4add76f
List files in the allowed folder.
無需輸入參數。
Permanently delete one file from the allowed folder. DESTRUCTIVE: the user must approve before anything happens. Describe the action to the user and wait for their answer.
path必填confirm_token已滿足適用檢查項的 90%
該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。
請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。
world.actions官方發布者
遠端 MCP 伺服器,讓助理在 actions.world 資料集上執行影片操作並追蹤執行結果。

org.aimlds
Meeting decisions and actions extract — Atom 的託管 MCP 服務,由 aimlds.org 發布;需要憑證;Streamable HTTP。

由 GitHub 帳號 redpop 發布
僅限 macOS 的搭配伺服器,讓助理在 Apple Mail 中把郵件移到其他信箱,並可選擇把單一附件存到本機資料夾。

由 GitHub 帳號 datadmon 發布
託管於 Apify 的 MCP 伺服器,以 JSON 回傳印度 NSE 公司揭露資訊:公告、除息日、董事會會議、大宗與鉅額交易及評等。

由 GitHub 帳號 datadmon 發布
由 Apify 託管的 MCP 端點,以 JSON 回傳印度 BSE 公司揭露資訊:公告、除息日、董事會會議、業績日期與鉅額交易。
{ "mcpServers": { "safe-files": { "command": "npx", "args": ["-y", "mcp-safe-actions"], "env": { "FILES_DIR": "/path/to/folder" } } } }npm install mcp-safe-actionsimport { rmSync, statSync } from "node:fs";import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";import { z } from "zod";import { registerConfirmedTool, fileAudit } from "mcp-safe-actions";
const server = new McpServer({ name: "files", version: "1.0.0" });
registerConfirmedTool(server, { name: "delete_file", description: "Permanently delete one file.", inputSchema: { path: z.string() },
// What WILL happen. Must not change anything. Be specific. preview: ({ path }) => ({ summary: `Permanently delete "${path}" (${statSync(path).size} bytes)`, details: ["This cannot be undone."], }),
// Runs only after the user approved. Re-check anything that could have changed. execute: ({ path }) => { rmSync(path); return { content: [{ type: "text", text: `Deleted ${path}` }] }; },
audit: fileAudit("./audit.jsonl"), // optional});
await server.connect(new StdioServerTransport());git clone https://github.com/techsaurabh-394/mcp-safe-actions.gitcd mcp-safe-actionsnpm installnpm run demo # prints both approval flowsnpm test # 101 tests{ "mcpServers": { "files": { "command": "node", "args": ["/absolute/path/to/mcp-safe-actions/dist/examples/files-server.js"], "env": { "FILES_DIR": "/absolute/path/to/test-folder", "AUDIT_LOG": "/absolute/path/to/audit.jsonl" } } }}sequenceDiagram participant M as AI model participant S as Your MCP server participant H as Human (client UI) M->>S: delete_file("report.pdf") S->>S: preview() builds the plan, nothing changes S->>H: Approval prompt: delete report.pdf? H-->>S: Approve or decline alt Approved S->>S: execute() and audit log S-->>M: Deleted report.pdf else Declined or no answer S-->>M: User did not approve. Nothing changed. endsequenceDiagram participant M as AI model participant S as Your MCP server participant H as Human (in chat) M->>S: delete_file("report.pdf") S-->>M: Nothing changed yet. Plan + one-time token M->>H: Shows plan, asks to go ahead H-->>M: Yes M->>S: delete_file("report.pdf", confirm_token) S->>S: token valid, unused, same arguments, not expired S->>S: execute() and audit log S-->>M: Deleted report.pdfregisterConfirmedTool(server, { name, title?, description, inputSchema, // like server.registerTool preview(args) => { summary, details? }, // no side effects execute(args) => CallToolResult, // after approval risk?: "destructive" | "write", // default "destructive" (sets annotations) mode?: "auto" | "elicit" | "token", // default "auto" approvalTimeoutMs?: number, // default 5 minutes; a timeout means "not approved" approver?: (req) => boolean | Promise<boolean>, // required for mode "channel" resolve?(args) => { label?, role?, url?, ... }, // optional: look up what the call actually targets, before the policy runs pinFields?: string[], // optional: which resolved fields must be unchanged right before execute (default: all) gate?: { pattern: RegExp, fields?: string[] }, // optional: if the resolved context matches, autoApprove is skipped autoApprove?(args, plan, context) => boolean, // optional: exactly `true` skips the prompt for low-risk calls store?: new ApprovalStore({ ttlMs, maxPending }), audit?: (event) => void | fileAudit(path), annotations?, // override tool annotations});registerConfirmedTool(server, { name: "click", inputSchema: { element_id: z.string() }, /* preview, execute ... */ resolve: ({ element_id }) => screen.describe(element_id), // { label: "Buy now", role: "button", url } gate: { pattern: /\b(buy|pay|delete|send|grant|allow)\b/i, fields: ["label"] }, autoApprove: (_args, _plan, ctx) => ctx?.role === "button", // only reached if the gate did not match});flowchart LR A[Push or pull request] --> B[CI] B --> B1[Build and 101 tests on Node 18, 20, 22] B --> B2[Demo runs end to end] B --> B3[Coverage report] B --> B4[Mutation check: 42 deliberate bugs] B --> B5[SVG diagrams valid] C[Tag v*] --> D[Release] D --> D1[Tests] D1 --> D2[GitHub Release with notes] D2 --> D3[npm publish with provenance, only if enabled]# .github/workflows/ci.ymlname: CIon: [push, pull_request]jobs: test: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - uses: actions/setup-node@v4 with: { node-version: 22, cache: npm } - run: npm ci - run: npm test # include a test that the dangerous tool does NOT run without approval