Allows creating and updating bookmarks in Pinboard by providing a URL, with automatic generation of a Japanese summary and tags, saved as private with overwrite support.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@pinboard-mcp 以下の記事をブックマークして https://example.com/ai-trends"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Claude Desktop から Pinboard にブックマークを保存するローカル MCP サーバー。
URL を伝えるだけで、Claude が日本語要約とタグを生成して Pinboard に保存する。
macOS（Claude Desktop がインストールされていること）
Node.js 24 以降
pnpm 9 以降
Pinboard の API トークン（設定画面から取得）
Related MCP server: Hatena Blog MCP Server
# 依存パッケージをインストール
pnpm install
# ビルド
pnpm build~/Library/Application Support/Claude/claude_desktop_config.json の mcpServers に以下を追加する。
{
"mcpServers": {
"pinboard-mcp": {
"command": "node",
"args": ["/path/to/pinboard-mcp/dist/index.js"],
"env": {
"PINBOARD_TOKEN": "username:xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
}
}
}
}/path/to/pinboard-mcp はこのリポジトリの絶対パスに置き換える。PINBOARD_TOKEN は username:token 形式（Pinboard 設定画面で確認できる値）をそのまま設定する。
設定を保存後、Claude Desktop を再起動する。
Claude Desktop で次のように伝えるだけでブックマークを保存できる。
https://example.comをブックマークして
Claude がページを取得・要約し、ブックマーク内容を提示してから確認を求める。「はい」と答えると Pinboard に保存される。
保存されるデータ：
項目 | 内容 |
URL | 指定した URL |
タイトル |
|
要約 | 140 字以内の日本語要約 |
タグ | 英小文字スラッグ形式のタグ（1〜5 個） |
すべてのブックマークは非公開（shared=no）で保存される。同一 URL が既に保存済みの場合は上書き更新される。
# テストを実行
pnpm test
# 型チェック・Lint・テストをまとめて実行
pnpm check
# Lint と自動整形
pnpm lint:write役割 | 採用技術 |
言語 | TypeScript 7 |
ランタイム | Node.js 24 |
MCP SDK |
|
バリデーション | zod |
テスト | Vitest |
Linter / Formatter | Biome |
選定根拠は docs/spec/003_architecture.md 参照。
指定した URL を Pinboard にブックマーク保存する。【重要】このツールを呼び出す前に、必ず以下の手順を踏むこと：(1) 記事の主張・背景・具体的な手法・結論・コード例などを網羅した日本語の詳細解説をユーザーに提示する。記事を読まなくても内容を把握できるレベルの情報量を目指し、文字数は制限しない。(2)「ブックマークしますか？」と確認を取る。(3) ユーザーが承認した場合のみ呼び出す。ツールに渡す summary は Pinboard 保存用の 140 コードポイント以内に凝縮した日本語要約であり、ユーザーへの提示内容とは別物にすること。tags は英小文字スラッグ（例: typescript, web-dev）で 1〜5 個とする。
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | ブックマーク対象の URL | |
| tags | Yes | 英小文字スラッグ形式のタグ。1〜5 個 | |
| title | Yes | ページタイトル | |
| summary | Yes | 140 コードポイント以内の日本語要約 |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure, and it discloses the critical trait: this is an external write to Pinboard gated by mandatory user confirmation. It also clarifies that the summary is a distinct Pinboard-oriented artifact, not the user-facing explanation. It omits minor behaviors like duplicate-URL handling, auth needs, and failure modes, but the essential side-effect safety information is present.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is long, but every sentence earns its place: the purpose is front-loaded, and the critical preconditions are flagged with 【重要】. The structure separates the mandatory workflow from parameter guidance, making the density scannable rather than bloated.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 4-param tool with no output schema and no annotations, the description supplies the essential operational context: the prerequisite explanation, the confirmation gate, and the summary/tags constraints. Missing items (duplicate bookmark behavior, return value, error cases) are minor and do not impair correct invocation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% so the baseline is 3, but the description adds real value beyond the schema. It specifies that summary must be a condensed Pinboard-saving text separate from what was presented to the user, and reinforces the tags format with concrete examples (typescript, web-dev). This communicates parameter intent, not just shape.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The opening sentence '指定した URL を Pinboard にブックマーク保存する' states a specific verb (save), resource (Pinboard bookmark), and target (URL). The purpose is unambiguous and immediately recognizable even before reading the lengthy workflow instructions. No siblings exist, so no differentiation is needed.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description mandates an explicit three-step precondition: present a comprehensive Japanese explanation of the article, ask 'ブックマークしますか？' for confirmation, and call the tool only upon user approval. This is unusually explicit about when to invoke the tool and what gates its use, leaving nothing to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0save_bookmarkScored across 1 tool
With only one tool, there is no possibility of confusion. save_bookmark has a single, clear purpose.
The single tool name follows a clear verb_object convention. There are no other tools to introduce naming inconsistency.
One tool is a thin but not absurd surface for a bookmarking service. It can support a narrow save-only workflow but feels limited for a full Pinboard MCP.
The Pinboard domain normally includes listing, updating, deleting, and searching bookmarks and tags. With only save_bookmark, users can create bookmarks but not retrieve or manage them, creating significant dead ends.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Save and organize web finds in persistent, user-controlled collections for AI assistants.
Persistent context for Claude. Your AI always knows your projects and next actions across sessions.
Publish pages straight from Claude as private, branded, tracked links.
Turn the LinkedIn posts you save into a knowledge base your assistant can search and cite.
