AIクライアントが、単一のアイテムまたは最大1000行のバッチに対して、校正されたYes/Noの判定、多肢選択、または数値スコアを取得できるようにします。
別名 foreverworks/mcps, @foreverworks/mcp-typesafe
- スター
- 5
- ツール
- 3
- ダウンロード数(30日間)
- 678
- 最新バージョン
- 0.1.3
- 最終コミット
- 2026年11月16日
- ライセンス
- MIT
TypeSafe MCPとは
AIクライアントが、単一のアイテムまたは最大1000行のバッチに対して、校正されたYes/Noの判定、多肢選択、または数値スコアを取得できるようにします。 3個のツールを提供しており、ask, ask_batch, healthなどがあります。
プロジェクトのREADMEから生成された概要
TypeSafe MCP server について
TypeSafeは、Jevモデルを通じて、AIアシスタントが校正された型付きの判定を取得するためのインターフェースを提供します。Yes/Noの確率、最大255の選択肢からの多肢選択、またはカスタムスケールによる数値スコアをリクエストできます。このMCP serverは、AIに多段階の推論やテキスト生成を行わせることなく、トランザクション、メール、レビューなどの大規模なデータセットを仕分けする必要がある開発者向けに設計されています。
ユーザーは、askツールを使用して単一のアイテムを処理するか、ask_batchを使用して最大1000行のバッチを一度に処理できます。このサーバーはJSONLファイルからのデータ読み取りをサポートしており、完全な結果をoutboxディレクトリに自動的に保存できます。使用するには、TypeSafeアカウントと、TYPESAFE_API_KEY_FILE環境変数経由でAPIキーを提供する必要があります。これは読み取り専用のサービスであり、コストは入力トークンによって決定され、レスポンスは通常約0.5秒で返されます。
できること
- 1トランザクションの一括分類AIを使用して、数百件のクレジットカード取引を食料品や交通費などの特定のカテゴリに分類できます。
- 2サポート通信の仕分け開発者はこのサーバーを使用して、受信メールの緊急度を判断し、適切な部署に割り当てることができます。
- 3高リスク顧客の特定顧客レビューを評価して、解約の可能性が最も高い特定の個人を見つけ出すことができます。
- 4特定のデータ行のフィルタリング大規模なデータセットから、特定の基準や信頼度のしきい値を満たすアイテムのみを抽出できます。
プロジェクトの README から作成されています。すべてのユースケースは、その引用と照らし合わせて確認されています。
TypeSafe MCPのインストール方法
フォーマット検証日 2026年11月10日Cursor に追加Gooseに追加Claude CodeにTypeSafe MCPをインストール
$ claude mcp add --env 'TYPESAFE_API_KEY_FILE=<TYPESAFE_API_KEY_FILE>' typesafe -- npx -y @foreverworks/mcp-typesafeCursorにTypeSafe MCPをインストール
~/.cursor/mcp.json
{
"mcpServers": {
"typesafe": {
"command": "npx",
"args": [
"-y",
"@foreverworks/mcp-typesafe"
],
"env": {
"TYPESAFE_API_KEY_FILE": "<TYPESAFE_API_KEY_FILE>"
}
}
}
}VS CodeにTypeSafe MCPをインストール
.vscode/mcp.json
{
"servers": {
"typesafe": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@foreverworks/mcp-typesafe"
],
"env": {
"TYPESAFE_API_KEY_FILE": "<TYPESAFE_API_KEY_FILE>"
}
}
}
}VS Code CLIにTypeSafe MCPをインストール
$ code --add-mcp '{"name":"typesafe","type":"stdio","command":"npx","args":["-y","@foreverworks/mcp-typesafe"],"env":{"TYPESAFE_API_KEY_FILE":"<TYPESAFE_API_KEY_FILE>"}}'Claude DesktopにTypeSafe MCPをインストール
claude_desktop_config.json
{
"mcpServers": {
"typesafe": {
"command": "npx",
"args": [
"-y",
"@foreverworks/mcp-typesafe"
],
"env": {
"TYPESAFE_API_KEY_FILE": "<TYPESAFE_API_KEY_FILE>"
}
}
}
}CodexにTypeSafe MCPをインストール
$ codex mcp add typesafe --env 'TYPESAFE_API_KEY_FILE=<TYPESAFE_API_KEY_FILE>' -- npx -y @foreverworks/mcp-typesafeCodex config.tomlにTypeSafe MCPをインストール
~/.codex/config.toml
[mcp_servers.typesafe] command = "npx" args = ["-y", "@foreverworks/mcp-typesafe"] [mcp_servers.typesafe.env] TYPESAFE_API_KEY_FILE = "<TYPESAFE_API_KEY_FILE>"
Gemini CLIにTypeSafe MCPをインストール
$ gemini mcp add -s user -e 'TYPESAFE_API_KEY_FILE=<TYPESAFE_API_KEY_FILE>' typesafe npx -- -y @foreverworks/mcp-typesafeGemini settings.jsonにTypeSafe MCPをインストール
~/.gemini/settings.json
{
"mcpServers": {
"typesafe": {
"command": "npx",
"args": [
"-y",
"@foreverworks/mcp-typesafe"
],
"env": {
"TYPESAFE_API_KEY_FILE": "<TYPESAFE_API_KEY_FILE>"
}
}
}
}HostedMCP が 2026年11月16日 に起動を確認
@foreverworks/mcp-typesafe 0.1.3 をネットワークもキーもない隔離コンテナでこのコマンドで起動しました。応答があり、3個のツールが一覧表示されました。
TypeSafe MCPツール 3
tools/list でサーバーから取得ask読み取り専用Ask TypeSafe's Jev, a System One model, typed questions about one state and get calibrated answers: a `noul` is the probability that a yes/no statement holds, a `choice` picks one of your options with a probability each, a `score` places the state on your ordered levels. Jev writes no text and does no multi-step reasoning: use it for fast, narrow judgments (routing, triage, relevance, a policy check) whose probabilities you then threshold. One narrow judgment per question; a no-match option where nothing may fit. A probability near 0.5 means undecided, not medium. About 0.5 s a call.
ask_batch読み取り専用The same questions over up to 1000 states, several requests at a time: triage postings, transactions, emails or passages without reading them all yourself. Replies with a summary per question (choice: counts; noul: mean and how many are >= 0.7 and <= 0.3; score: mean, min, max) and each row's compact answers (a noul's probability, a choice's pick and confidence, a score and its confidence). For a large batch pass list 'none' or 'failed' with save true: every row's full answers go to a JSONL file in the outbox and the reply names it with its sha256 and size. A row that fails is reported and the rest go on. Rows already in a file go as items_file, not typed into items: that keeps them out of your context and exactly as written. thresholds turn each answer into act, review or abstain; where lists only the rows that meet conditions, top only the k highest by a question, and counts tallies the answers (and verdicts) of the rows where keeps: over a large batch they keep the reply to the rows that matter. A client that asks for progress hears the rows done; cancelling stops the rows not yet asked.
health読み取り専用Check that the TypeSafe API key is present and accepted, and list the models it can use. Run this before concluding the MCP is broken.
| ask | Ask TypeSafe's Jev, a System One model, typed questions about one state and get calibrated answers: a `noul` is the probability that a yes/no statement holds, a `choice` picks one of your options with a probability each, a `score` places the state on your ordered levels. Jev writes no text and does no multi-step reasoning: use it for fast, narrow judgments (routing, triage, relevance, a policy check) whose probabilities you then threshold. One narrow judgment per question; a no-match option where nothing may fit. A probability near 0.5 means undecided, not medium. About 0.5 s a call. | 読み取り専用 |
| ask_batch | The same questions over up to 1000 states, several requests at a time: triage postings, transactions, emails or passages without reading them all yourself. Replies with a summary per question (choice: counts; noul: mean and how many are >= 0.7 and <= 0.3; score: mean, min, max) and each row's compact answers (a noul's probability, a choice's pick and confidence, a score and its confidence). For a large batch pass list 'none' or 'failed' with save true: every row's full answers go to a JSONL file in the outbox and the reply names it with its sha256 and size. A row that fails is reported and the rest go on. Rows already in a file go as items_file, not typed into items: that keeps them out of your context and exactly as written. thresholds turn each answer into act, review or abstain; where lists only the rows that meet conditions, top only the k highest by a question, and counts tallies the answers (and verdicts) of the rows where keeps: over a large batch they keep the reply to the rows that matter. A client that asks for progress hears the rows done; cancelling stops the rows not yet asked. | 読み取り専用 |
| health | Check that the TypeSafe API key is present and accepted, and list the models it can use. Run this before concluding the MCP is broken. | 読み取り専用 |
設定
TYPESAFE_API_KEY_FILE必須Path of a file holding your TypeSafe API key.
| 名前 | 説明 | 種類 |
|---|---|---|
| TYPESAFE_API_KEY_FILE | Path of a file holding your TypeSafe API key. | 必須 |
自動チェック
実行日 2026年11月12日完全なレポート →自動チェックに合格しました。これはセキュリティ監査ではありません。コードを読み、サーバーに必要最小限のアクセス権を与えてください。 チェックの仕組み
- ソースリポジトリがリンクされており、アクセス可能合格
- オープンソースライセンス (OSI)合格
- 過去1年以内にコミットまたはリリースあり合格
- パッケージ名が有名なパッケージに酷似していない合格
- リモートエンドポイントがHTTPSを使用未チェック
- 破壊的なツールがラベル付けされている合格
- モデルへの隠れた指示がない合格
- パッケージおよびその依存関係に重大な勧告なし合格
README
GitHubで見るTypeSafe MCP server
Ask TypeSafe's Jev for typed, calibrated judgments from
your AI assistant: how likely a yes/no statement is to hold, which of your
options fits, or where something sits on your own scale. Put the same
questions to one state or to up to 1000 rows at once, so the assistant can
triage postings, transactions or emails without reading each one. Every tool
is a read: nothing changes anywhere, and no tool is money.
You need a TypeSafe account and an API key from its console. Calls are billed by input tokens, a fraction of a cent each, and Jev answers in about half a second.
What you can ask
- "Go through the 300 postings in postings.jsonl and list the remote senior roles, with how sure Jev is of each."
- "Is this support email urgent, and which team should take it: billing, technical or sales?"
- "Sort last month's 600 card transactions into groceries, transport, software or other, and count each."
- "Rate these 40 reviews on how likely the customer is to leave, and show me the ten most at risk."
- "Which of these passages answer 'how do I reset my password'? Flag the ones Jev is unsure about for me to read."
Install
With Claude Code:
claude mcp add typesafe -e TYPESAFE_API_KEY_FILE="<path to API key file>" -- npx -y @foreverworks/mcp-typesafe
Claude Desktop (Settings, Developer, Edit Config) and most other clients read the same command as JSON:
{
"mcpServers": {
"typesafe": {
"command": "npx",
"args": ["-y", "@foreverworks/mcp-typesafe"],
"env": {
"TYPESAFE_API_KEY_FILE": "<path to API key file>"
}
}
}
}
Each runs the published package on Node 22.19 or newer. VS Code and the Claude Desktop bundle ask for TYPESAFE_API_KEY_FILE as they add the server. Cursor reads it from your environment instead (${env:TYPESAFE_API_KEY_FILE}), so set it before you start Cursor. Configuration says what it holds.
From a clone
git clone https://github.com/foreverworks/mcps.git && cd mcps
npm ci && npm run build
claude mcp add typesafe -e TYPESAFE_API_KEY_FILE="<path to API key file>" -- node "$PWD/servers/typesafe/dist/index.js"
In a JSON configuration, the command is node and its one argument the full path of servers/typesafe/dist/index.js. To share one running copy between several clients over HTTP, see docs/deploying.md.
Configuration
| Variable | Required | Meaning |
|---|---|---|
TYPESAFE_API_KEY_FILE |
yes | File holding the API key. For a local run, TYPESAFE_API_KEY may hold the key itself instead. |
TYPESAFE_MODEL |
no | The model a call uses when it names none. Default jev-latest, which tracks the newest Jev. |
TYPESAFE_BASE_URL |
no | Default https://api.typesafe.ai. |
TYPESAFE_OUTBOX_DIR |
no | Where ask_batch with save writes its JSONL. Default outbox/ in the state folder. Files are written 0640, so a deployment can share the folder with the user's group. |
TYPESAFE_STATE_DIR |
no | The state folder; the default is in docs/configuration.md. |
Without a key the server still starts and lists its tools, and each call
answers a config error naming the variable. ask_batch reads an
items_file from MCP_FILE_ROOTS; that and the other variables every server
shares are in docs/configuration.md.
Getting an API key
- Sign in at https://console.typesafe.ai and open Keys.
- Create a key and write it to a file only you (or the server's user) can read.
- Point
TYPESAFE_API_KEY_FILEat that file.
Give each deployment its own key, so you can revoke one alone and tell them apart on the console's usage page. Only a key from TypeSafe's own console works; a reseller's key works only through the reseller's gateway.
First run
Ask your assistant to run TypeSafe's health check. It answers with the API
it talks to, the default model, and the models your key reaches. A missing key
shows up here as config, a refused one as auth, before any batch starts.
Tools
read changes nothing; every tool here is a read (billing by input tokens is metering, not a payment).
| Tool | Access | What it does |
|---|---|---|
health |
read | Whether the key is present and accepted, and which models it reaches. |
ask |
read | Every question about one state in one round trip; the answers as TypeSafe returns them, less a Score's legend. |
ask_batch |
read | The same questions over up to 1000 states, typed in or from a file, 8 in flight by default: a summary per question and compact rows, with save the full rows as JSONL in the outbox. Optional act, review or abstain verdicts from your thresholds, and where, top and counts to list only the rows that matter. |
A question is a noul (the probability that a yes/no statement holds), a
choice (one of 2 to 255 options, with a probability for each) or a score
(the state placed on 2 to 10 levels you order from lowest to highest).
Troubleshooting
config. No key: set TYPESAFE_API_KEY_FILE to a file holding a key from
console.typesafe.ai/keys, or TYPESAFE_API_KEY for a local run.
auth. TypeSafe refused the key. Check it still exists at
console.typesafe.ai/keys, and replace the file if it was revoked.
input. The question cannot be answered as asked; the message names the
field (a choice needs at least two options, a score 2 to 10 levels). Above 20
rows, every row needs an id.
rate_limited. TypeSafe is still throttling or overloaded after the
server's own retries. Wait a minute, or send the batch with a lower
concurrency.
api. Any other refusal, a 5xx that outlived its retries, or no connection
(status 0). Retry once; if it persists, run health and read the server's log.
drift. TypeSafe answered in a shape this server does not know (its SDK's
types may have moved on). Please report it.
How it works
The network. An evaluation has no side effect, so every request
(POST /v1/systemone, GET /v1/models) may repeat: a 5xx or a dropped
connection is tried again, and a 429 or 529 ("overloaded") waits as a rate
limit, as TypeSafe's SDK does. A call times out after a minute. Every answer
is checked against the SDK's types (@typesafe-ai/sdk, types only, in
src/domain/schema.ts).
Batches. A batch goes on past a failed row and reports it, but stops at a failure that would fail every row (no key, a refused key), so a thousand rows do not each earn a 401. Answers join their rows by id, required above 20 rows: a position is only safe to join on while you can check the rows by eye.
Rows from a file. Instead of typing rows into items, pass items_file:
JSONL with one {"id", "state"} object per line, or a JSON array of them, by
path inside MCP_FILE_ROOTS (or as content_base64 with its sha256, from a
script). The rows never pass through a model, so they cannot change on the
way; a malformed row fails the call with its line number before anything is
sent. The reply names the file's sha256 and size.
Verdicts. With thresholds for a question, each row's answer gets act
(sure enough to act on), review (a person should look) or abstain (do not
use it), from the answer's certainty: for a noul |p - 0.5| x 2 (p 0.95 and
p 0.05 are both 0.9), for a choice or a score the confidence TypeSafe returns.
The certainty is rounded down to three places, the number the row shows, and
must be strictly above act_above to act and above review_above to be
reviewed. Once thresholds are tuned, pin model to a versioned Jev:
jev-latest can move under them.
Choosing rows. where lists only the answered rows meeting every
condition you give per question: true or false for a noul (a yes is
p >= 0.7, a no p <= 0.3, as the summary counts them), labels for a choice,
min and max on the value a row shows, and verdict. top keeps the k
rows with the highest p or score for one question, and counts tallies the
answers of the rows where kept; the summary still covers every row. Every
condition is checked before a request is sent, so a label that is not an
option fails the call, not a thousand rows.
npx tsx scripts/smoke.ts drives the built server against the real API, for
a few thousand input tokens: health, one ask per question type, a
five-row batch saved to the outbox, and one input failure that never reaches
TypeSafe.
Limitations
- Judgments only. Jev writes no text and does no multi-step reasoning.
Keep rules, lookups and arithmetic in code, and threshold the probabilities
(
thresholdsdoes it per question). wherejoins conditions with "and" only. For "or", read the rows from a saved JSONL, or ask a question that states the alternatives.- At most 1000 rows a call; split a larger set.
- A batch's reply lists every row unless
listsays otherwise. For a large batch, passlist: "none"or"failed"withsave, and read the JSONL.
Other servers for TypeSafe
Searched 2026-10-15. TypeSafe publishes none: its GitHub organisation has the SDKs, agent skills and an n8n node. Two community servers are on npm:
typesafe-mcp: stdio under Bun, the key in an environment variable, batches over files matched by a glob, with an act, review or abstain verdict per answer.@y0usaf/typesafe-mcp: stdio, oneevaluatetool.
This one exists for what neither offers: it also runs over HTTP as a service
of its own, with its key in a file the calling client never reads; it saves a
batch's full answers as JSONL in an outbox and names the file's SHA-256; and
it follows the conventions every server in this repository shares. Its
verdicts and its where, top and counts follow typesafe-mcp's, with
thresholds set per question rather than per call.
TypeSafe MCPのよくある質問
TypeSafe MCPをClaude Codeに追加するには
ターミナルでclaude mcp add --env 'TYPESAFE_API_KEY_FILE=<TYPESAFE_API_KEY_FILE>' typesafe -- npx -y @foreverworks/mcp-typesafeを実行します。すべてのプロジェクトで利用するには--scope userを追加します。
TypeSafe MCPをCursorに追加するには
~/.cursor/mcp.jsonを開き、上記のインストールセクションに示すCursor設定を使ってmcpServersの下にサーバーを追加します。
TypeSafe MCPが提供するツールは
3個のツールを提供しており、ask, ask_batch, healthなどがあります。
TypeSafeからはどのような回答が得られますか?
「noul」(Yes/Noの表明に対する確率)、「choice」(2〜255の選択肢のいずれか)、または「score」(2〜10段階のスケールによる値)を取得できます。
一度に何行まで処理できますか?
1回のバッチリクエストで最大1000行まで処理可能です。
TypeSafeを使用するとデータが変更されますか?
いいえ、サーバーが提供するすべてのツールは読み取り専用の操作です。