Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者による説明(英語の原文)
Rank log patterns by volume and cost, then compact, tier down or offload the excess over each cap
この3つを1つのスコアにまとめることはしません。意見、測定、人気はそれぞれ別のものだからです。
このサーバーのやさしい解説はまだ書いていません。以下はレジストリから読み取った事実です。
使っているAIアプリを選んで、設定をコピーしてください。
LOG10X_API_KEY「秘密」と表示された値は、あなたのアカウントの鍵です。絶対に人に教えたり、チャットに貼り付けたりしないでください。
必要なもの
どのアプリを使っていますか?
上でアプリを選んでください。選んだアプリを覚えておくので、次からはボタンに「…にインストール」と表示されます。
Claude Desktop → Settings → Developer → Edit Config を開き、claude_desktop_config.json にこれを貼り付けます
{
"mcpServers": {
"log10x": {
"command": "npx",
"args": [
"-y",
"log10x-mcp"
],
"env": {
"LOG10X_API_KEY": "your-value"
}
}
}
}ターミナルでこれを実行します
claude mcp add log10x --env LOG10X_API_KEY=your-value -- npx -y log10x-mcp動かす前に、空欄の値を入力してください。
プロジェクトの .cursor/mcp.json、または ~/.cursor/mcp.json に貼り付けます
{
"mcpServers": {
"log10x": {
"command": "npx",
"args": [
"-y",
"log10x-mcp"
],
"env": {
"LOG10X_API_KEY": "your-value"
}
}
}
}ブラウザがアプリを開いてよいか確認し、そのあとアプリが設定を表示して確認を求めます。キーの値は空欄のままなので、自分で入力してください。
プロジェクトの .vscode/mcp.json に貼り付けます
{
"servers": {
"log10x": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"log10x-mcp"
],
"env": {
"LOG10X_API_KEY": "your-value"
}
}
}
}ブラウザがアプリを開いてよいか確認し、そのあとアプリが設定を表示して確認を求めます。キーの値は空欄のままなので、自分で入力してください。
これはパソコンで動くサーバーですが、ChatGPTがつなげるのはインターネット上のサーバーだけです。ほかのアプリを選んでください。
選んだアプリは、この端末のCookieに保存されます。アカウントは不要です。
この設定は、作者のレジストリ登録情報から作成しています。READMEと違う場合は、READMEに従ってください。
パソコンで動くサーバーです。インストールすると、ツール一覧がAIアプリに表示されます。
誰でも見られる情報をチェックしており、コードは読んでいません。当てはまらない項目は数に入れていません。 · 2026年11月7日にチェック
レジストリ名は、リポジトリを所有するGitHubアカウント log-10x で確認済みです
このリポジトリはまだGitHubから読み取っていません
このリポジトリはまだGitHubから読み取っていません
これはパソコンで動くため、インターネット上のエンドポイントはありません
パソコンで動き、代わりにローカルの設定を使います
パソコンで動くサーバーです。ツール一覧を読み取れるのは、インターネット上のエンドポイントがある場合だけです
レジストリでこのカテゴリにある有効なすべてのサーバーと比較しています。スターは、公開者がリポジトリの所有者である場合だけ数えます。
当てはまる基本チェック1項目のうち1項目に合格しました。コードは読んでいません。キーやトークンを追加する前に、サーバーがどんなアクセスを求めているかを確認し、必要な権限だけを持つキーを使ってください。
入力が必要なもの:LOG10X_API_KEY
あなたのパソコンで動きます。必要なときにAIアプリが起動し、データは作者のサーバーを通りません(サーバー自体がオンラインサービスを呼び出す場合を除きます)。
GitHubリポジトリの最終更新は–、最新バージョン(1.30.99)がレジストリに届いたのは一昨日です。
Claude Desktop、Claude Code、Cursor、VS CodeなどのMCP対応アプリで使えます。パソコンにNode.jsも必要です。
同じカテゴリには:Spec Driven Development, Remote Desktop Commander, Excaliself
同じカテゴリで、ほかの作者が作ったサーバー。スターの多い順。
カードの見方:★ GitHubスターi3つの柱:レビュー · チェック · 利用状況i更新=最新バージョンをレジストリに公開した日i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Find, read, create, edit, and organize Excalidraw drawings in your self-hosted Excaliself instance.
作者のリポジトリから書かれたとおりにコピーしたものです。PopNx MCPは内容を確認・推奨していません。HTML、画像、スクリプトは取り除いています。リンクはこのサイトの外へ移動します。
Ask your AI assistant about log costs, log patterns, and incidents in plain English, and get structured answers backed by per-pattern metrics. "Why did our log costs spike this week?", "what's driving payments-svc volume?", "pull the payment_retry events for acme-corp from Jan 15 to Apr 15." The server exposes the 10x platform to Claude and any MCP-compatible assistant as a set of tools.
10x groups your logs by message type. The same lines repeat over and over with only the timestamp or request ID changing, so 10x collapses each flood of near-identical lines into one message type and ranks them by volume and cost. Names come from symbols compiled from code, so the same line gets the same name on every node under the same engine, library and configuration.
This server is open source under MIT. A Log10x product license (API key) unlocks the account-scoped tools; without one, the server boots read-only against a public demo dataset so the tools can be explored immediately.
| You say | What happens |
|---|---|
| "Where do I start? Help me cut log costs." | A guided menu asks what you want (cut cost, investigate, install) and routes to the right next step, plus what's new and what changed. log10x_start, log10x_whats_new, log10x_whats_changing |
| "Sign me in." / "Who am I connected as?" | GitHub sign-in mints and stores an API key, and shows the environments your account can reach. log10x_signin_start, log10x_login_status, log10x_create_env |
| "Show me how to install this on my stack." | Paste-ready setup steps for your own pipeline, fetch-back wiring, and a recommended action per service. A policy ships as a pull request to merge, or, for a ConfigMap, is written only by a second, explicit call. log10x_advise_install, log10x_advise_retriever, log10x_configure_engine |
| "Estimate savings on my data before I deploy." | A no-install savings report from a local log file or your existing log platform. log10x_poc_from_local, log10x_poc_from_siem_submit, log10x_poc_from_siem_status. A keyless boot attaches to the public demo dataset and registers these three as well, since they run on the user's machine and never touch the demo backend. LOG10X_DEMO_FALLBACK=off boots into POC mode instead, which drops log10x_top_patterns, log10x_investigate, log10x_savings and every Retriever tool. |
| "How much can I cut, and how much have I cut?" | Projected and realized savings with the per-pattern math behind every number. log10x_estimate_savings, log10x_savings, log10x_commitment_report |
| "What's driving my cost right now?" | The repeating message types and services driving volume and cost, and what moved week over week. log10x_top_patterns, log10x_whats_changing, log10x_services |
| "Why did payments-svc spike?" | A single-call investigation: timeline, correlated patterns, and the strongest temporal evidence, with confidence shown so nothing is presented as proven cause. log10x_investigate, log10x_metrics_that_moved, log10x_metric_overlay |
| "This message type is noise, cut it." | Sample, drop, compact, tier down, or offload a chosen message type, with examples and trend so the action is clear. log10x_pattern_mitigate, log10x_pattern_examples, log10x_pattern_trend |
| "We are 100% Lambda, no cluster. Install there." | The install plan comes back for the serverless estate: the engine as a Lambda extension paired with the function's OTel collector, a CDK construct, and an AWS-native recurring tick over an S3 config plane. log10x_advise_install, log10x_setup_recurring |
| "Get me the actual events I offloaded." | Fetch the exact events back on demand from your own S3. log10x_retriever_query, log10x_offload_add, log10x_retriever_register |
| "Recognize more of my message types." | Compile your own source code, binaries and container images into a symbol library so the engine names more of your message types. log10x_compile, log10x_compile_link, log10x_compile_status |
Claude Code
claude mcp add --transport stdio --env LOG10X_API_KEY=your-api-key log10x -- npx -y log10x-mcp
Verify with /mcp. Omit the --env flag to run read-only against the public demo dataset.
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": { "log10x": { "command": "npx", "args": ["-y", "log10x-mcp"], "env": { "LOG10X_API_KEY": "your-api-key" } } } }
Cursor, Windsurf, and other MCP clients
Add an mcpServers entry of the same shape: "command": "npx", "args": ["-y", "log10x-mcp"], and LOG10X_API_KEY in env. Omit the key to run read-only demo mode.
Full setup walkthrough: https://doc.log10x.com/apps/mcp/
Chart images (include_chart on log10x_top_patterns and log10x_pattern_trend) need chart.js and chartjs-node-canvas, optional peers that are not installed with the package; without them charts render as text.
The Elasticsearch, Azure Monitor and GCP Cloud Logging connectors, and the Azure App Configuration and GCP Secret Manager config stores, also use optional SDKs that keep the default install small. Add them when the server starts, for example npx -y -p log10x-mcp -p @azure/monitor-query -p @azure/identity log10x-mcp; the connector names the packages it needs if they are missing.
Account-scoped tools need a Log10x API key. Two ways to get one:
log10x_signin_start, complete the flow, then log10x_signin_complete. The key is stored for you. log10x_login_status shows who you're connected as.LOG10X_API_KEY directly.No key? The server boots read-only against the public 10x demo dataset (the same sample data the website console shows), so analysis, discovery, and the install advisors all work out of the box. Mutating tools are disabled in this mode so shared demo state stays untouched. Set LOG10X_DEMO_FALLBACK=off to boot into pure POC mode (savings estimates on your own files, no account). The mode is fixed at startup; restart the server to re-detect.
All optional. The common path is just LOG10X_API_KEY (or no key at all).
| Variable | Purpose |
|---|---|
LOG10X_API_KEY | Your Log10x API key. Omit to boot read-only on the public demo dataset. |
LOG10X_API_BASE | API base URL override (default https://prometheus.log10x.com). |
LOG10X_ENV_ID | Pin a specific environment id instead of autodiscovering from your account. |
LOG10X_DEMO_FALLBACK | Set to off to boot into POC mode instead of attaching to the public demo dataset when no key is set. |
LOG10X_MCP_READ_ONLY | Set to true or 1 to force read-only mode (all mutating tools refuse). |
LOG10X_CUSTOMER_METRICS_URL / _TYPE / _AUTH | Point at your own metrics backend (Prometheus, Grafana Cloud, Mimir, Thanos, AMP) for BYO-metrics setups. |
DD_API_KEY / DD_APP_KEY / DD_SITE | Datadog credentials for backfilling per-pattern metrics into Datadog. |
PROMETHEUS_REMOTE_WRITE_URL / PROMETHEUS_URL | Prometheus endpoints for backfilling and reading per-pattern series. |
__SAVE_LOG10X_RETRIEVER_URL__ / __SAVE_LOG10X_RETRIEVER_BUCKET__ | Retriever query endpoint and the bucket it indexes, for fetch-back. Both are needed; log10x_retriever_register sets the same pair. A Lambda function URL is signed with your AWS credentials. |
__SAVE_LOG10X_RETRIEVER_TARGET__ | Default target app prefix for fetch-back queries, e.g. app. |
LOG10X_RETRIEVER_LOG_GROUP | CloudWatch log group the Retriever writes query events to, the value of its TENX_QUERY_LOG_GROUP. With it, a zero-result query says which stage matched nothing. |
LOG10X_RETRIEVER_TIMEOUT_MS / LOG10X_RETRIEVER_POLL_MS | Fetch-back query timeout (default 90000) and poll interval (default 1500). |
LOG10X_RETRIEVER_AUTH_HEADER / LOG10X_RETRIEVER_AUTH_VALUE | Override the fetch-back auth header (defaults derive from the active environment). |
LOG10X_OFFLOAD_BUCKET / LOG10X_STREAMER_BUCKET | S3 bucket names the offload tools manage. |
TENX_LICENSE_KEY | Engine license key. Passed through to the engine on both the compile path and the local-engine run path (log10x_resolve_batch, log10x_extract_templates), in docker mode as well as local mode. In docker mode, a key the engine rejects is withheld on an automatic retry so the run falls back to the image's built-in limited license rather than failing; the downgrade is logged on the server's stderr. |
OTEL_EXPORTER_OTLP_ENDPOINT | Sends the server's own tool-call traces to an OTLP endpoint. The OpenTelemetry SDK is an optional peer and is not installed with the package: install @opentelemetry/sdk-node, @opentelemetry/exporter-trace-otlp-http and @opentelemetry/resources beside it, or tracing stays off and the server says so on stderr. |
TENX_LICENSE_FILE | Path to an engine license file, the alternative to TENX_LICENSE_KEY. Read by the engine itself, so it must name a path the engine process can see (inside the container when the engine runs in docker). |
LOG10X_TENX_MODE | local or docker for the local-engine and compile tools. Unset auto-detects: a tenx on PATH wins, docker is the fallback. Set docker to run the engine image instead of a local install, which is also the escape when a local tenx has no license. |
LOG10X_TENX_PATH | Path to the local tenx binary (used when the resolved mode is local). Defaults to tenx on PATH. |
LOG10X_RUNTIME_IMAGE | Engine image for the run path in docker mode (default log10x/pipeline-10x:latest). Accepts the alias native, which selects the GraalVM-native log10x/edge-10x. |
LOG10X_TENX_IMAGE | Engine image shared by the run path and, as a fallback, log10x_compile. A runtime-flavor image here is refused on the compile path. |
LOG10X_COMPILER_IMAGE | Compiler image for log10x_compile / log10x_compile_link (default log10x/compiler-10x:1.1.136 pinned by digest, so every run goes through the same compiler; set it to :latest to track releases; falls back to LOG10X_TENX_IMAGE when unset). Only the compiler flavor can build a symbol library. |
LOG10X_GH_REPO / LOG10X_GITOPS_REPO_PATH | GitHub repo and local path for GitOps-aware config tools. |
10x keeps logs queryable without paying to index everything. The Receiver ships budget-overflow noise to your own S3, and the Retriever fetches those exact events back on demand when you need them. log10x_offload_add and log10x_retriever_register wire up the offload-and-fetch path; log10x_backfill_metric pushes per-pattern volume and cost series into your existing Datadog or Prometheus so the cost views light up alongside the rest of your dashboards.
The server runs locally as a subprocess of your AI client. Only pre-aggregated metric queries (per-pattern volume and cost) leave the machine, never raw log content. Local triage with log10x_resolve_batch runs entirely on your machine over stdin and stdout. Connections to your own log platform use read-only credentials. Set LOG10X_MCP_READ_ONLY=true to refuse every mutating tool regardless of mode.
For an evaluation that has to be checkable rather than promised: MCP server and log engine in one container started with --network none, over a log sample you exported yourself with your own credentials.
docker run -i --rm --pull=never --network none --hostname localhost \
--cap-drop ALL --security-opt no-new-privileges \ -v "$PWD/poc/logs:/data:ro" -v "$PWD/poc/out:/out" \ -e TENX_LICENSE_KEY -e TENX_AIRGAPPED=true log10x/poc:local
There is no npx variant of this line, and that is not an oversight. npx resolves and executes a packag
冒頭だけを表示しています。続きはリポジトリで読めます