Agentum Entity Resolver
Company registry, VAT, sanctions, vendor KYB, IBAN and wallet checks; pay per call by x402 or card.
作者による説明(英語の原文)
Chemical safety intelligence — 25 tools: compatibility, hazards, PPE, compliance, SDS lookup
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。
MSDS_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": {
"msds-chain": {
"command": "npx",
"args": [
"-y",
"msds-chain-mcp"
],
"env": {
"MSDS_API_KEY": "your-value"
}
}
}
}ターミナルでこれを実行します
claude mcp add msds-chain --env MSDS_API_KEY=your-value -- npx -y msds-chain-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": {
"msds-chain": {
"command": "npx",
"args": [
"-y",
"msds-chain-mcp"
],
"env": {
"MSDS_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": {
"msds-chain": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"msds-chain-mcp"
],
"env": {
"MSDS_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 littleblakew 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:MSDS_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(1.5.12)がレジストリに届いたのは4 日前です。
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:Agentum Entity Resolver, Contract Compass, Agent Data Tools: Company, compliance and checks
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
Company registry, VAT, sanctions, vendor KYB, IBAN and wallet checks; pay per call by x402 or card.
Korean public procurement law: rule-engine rulings, statutes search, live court precedents
38 tools: OFAC sanctions, counterparty checks, company, domain, email. 25 free calls.
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。
Chemical safety intelligence for AI-assisted experiment design.
An MCP server that gives AI agents (Claude, ChatGPT, Claude Code, Gemini CLI, Copilot Studio, and any other MCP client) 27 tools for chemical safety reasoning: compatibility checks, hazard analysis, regulatory compliance, PPE, storage, transport, and signed audit reports. Every answer names the supplier SDS it came from, with a link to the original PDF.
MSDS Chain, powered by ChainSDS: a verified, always-current chemical safety database.
Built for researchers who design experiments with AI and want safety verification inside the workflow rather than as a separate step: whether the chemicals on a deck are compatible, which mixing orders are dangerous, what PPE a specific handling step needs, how a compound is treated under EU REACH or US OSHA/TSCA, and what a GLP/GMP reviewer will ask for afterwards.
Hosted endpoint. Nothing to install, nothing to run locally:
claude mcp add msds-chain --transport http https://mcp.lagentbot.com/mcp
Then run /mcp in Claude Code and sign in. A browser opens, you enter your email, and you paste back the code we send you. Your account is provisioned automatically and the free plan includes a credit allowance. There is no API key to create beforehand.
On another platform? See Platform Setup below.
| Tool | Description |
|---|---|
| `batch_safety_check` | One-call report for a chemical list: pairwise compatibility + per-chemical risk warnings |
| `check_regulatory_lists` | Cross-reference a chemical against 23 regulatory watch lists across 8 jurisdictions + 3 international conventions |
| `review_synthesis_route` | Screen a multi-step synthesis route for incompatibilities within each step and between carried-over residue and the next step; never declares a route safe |
| `assess_thermal_hazard` | Thermal hazard screening: energetic-group and oxygen-balance flags, plus a Stoessel class from process values you supply |
| `get_sds_section` | Retrieve a specific SDS section (1-16) for a chemical |
| `get_sds_document` | Signed download URL (~5 min) for the original SDS/MSDS PDF; includes source provenance |
| `get_chemical_alternatives` | Safer substitutes for restricted or high-risk chemicals |
| `validate_protocol_chemicals` | Extract & validate chemical names from protocol text or code |
| `check_mixing_order` | Safe addition sequence for reagent pairs (e.g., acid into water) |
| `get_waste_disposal` | Waste classification, container type, and disposal procedures |
| `upload_msds_pdf` | Upload MSDS PDF for AI-powered parsing and data extraction (requires API key) |
| `validate_sds` | Check an SDS you uploaded for contradictions between its own sections, gaps, age and US OSHA HazCom format; every finding quotes the lines it is based on |
| `draft_sds_sections` | Draft Sections 8 and 15 of an SDS for your own mixture from its ingredient list; every line cites its source, and missing ingredients are listed rather than guessed |
| `compare_sds_versions` | Hazard-change diff between two SDS versions (H-code additions/removals + whether they change a verdict) |
check_chemical_compatibility | Pairwise compatibility for 2+ chemicals |
get_chemical_risk_warnings | GHS classification, H-codes, signal words, flash point |
get_ppe_recommendation | Gloves, eye protection, respiratory, body protection |
get_storage_guidance | Storage class, cabinet type, temperature, isolation rules |
get_emergency_response | Spill, fire, or exposure emergency procedures |
get_exposure_limits | OEL/TLV/PEL/MAC across US, EU, JP, CN, INT |
get_transport_classification | UN number, hazard class, packing group, ADR/IATA/IMDG |
check_regulatory_compliance | Multi-region compliance status, list-backed for EU, US, CN, JP, CA, AU, SG |
search_chemical_database | Look up chemicals by name, synonym, or CAS number |
ask_chemical_safety | Natural language catch-all for any safety question |
create_audit_session | Full audit with signed PDF report (requires API key) |
get_audit_report | Download link for the signed audit PDF |
Experiment protocol review:
User: I'm planning a Grignard reaction with magnesium turnings, diethyl ether,
and bromobenzene. Check if this setup is safe.
Agent: → calls batch_safety_check(["magnesium", "diethyl ether", "bromobenzene"]) → Returns: compatibility matrix + per-chemical risk warnings (with source SDS)
Signed audit report:
User: Create a safety audit report for our quarterly review.
Chemicals: acetone, methanol, ethanol, isopropanol, hexane.
Agent: → calls create_audit_session("Q2 2026 Solvent Cabinet Review", [...]) → calls get_audit_report("SESSION-ID") → Returns: signed PDF URL (Ed25519 signature, suitable for GLP/GMP compliance)
One hosted server, two transports. Pick the row your client supports:
| Transport | Endpoint | Notes |
|---|---|---|
| Streamable HTTP | https://mcp.lagentbot.com/mcp | Preferred: broadest and most robust |
| SSE | https://mcp.lagentbot.com/sse | For clients that only speak SSE |
Authentication is either OAuth (the client opens a browser sign-in on first connect) or a static header, Authorization: Bearer sk-msds-your-key, for headless clients. Create a key at msdschain.lagentbot.com > API Keys.
Settings > Connectors, search msds-chain, and enable it. Sign-in happens on first use.
Three routes, any one is enough:
# 1. Remote server only
claude mcp add msds-chain --transport http https://mcp.lagentbot.com/mcp
/plugin install msds-chain@claude-plugins-official
claude mcp add msds-chain -- npx -y msds-chain-mcp@latest
Manual config in ~/.claude.json:
{
"mcpServers": { "msds-chain": { "type": "http", "url": "https://mcp.lagentbot.com/mcp" } } }
Routes 2 and 3 of the plugin also install the /msds-safety-check skill: it spots chemicals in your conversation and offers a check, and runs a guided audit for lab protocols or EHS compliance.
Listed in the OpenAI Plugin Directory. Search MSDS Chain in ChatGPT, enable it, and sign in when prompted. Works on web, desktop, and mobile.
Listed in the Gemini CLI extensions gallery:
gemini extensions install https://github.com/littleblakew/msds-chain-mcp
Gemini asks you to confirm twice: once to trust the workspace, once to review the extension and the skill it carries. Your first chemical question starts the sign-in automatically; the server supports Dynamic Client Registration, so Gemini registers itself and completes the OAuth exchange on its own. Check the connection with gemini mcp list, then just ask in chat:
Can acetone and sodium hypochlorite be stored in the same cabinet?
codex marketplace add https://github.com/littleblakew/msds-chain-mcp
MSDS Chain ships a certified Power Platform connector, so there is nothing to configure by hand. In Copilot Studio, add a tool from the connector directory and search MSDS Chain; the connector handles the OAuth exchange against the same hosted endpoint.
These clients do not speak streamable HTTP yet, so point them at /sse and pass a static key:
https://mcp.lagentbot.com/sseAuthorization : Bearer sk-msds-your-keyUse the endpoint table above. Clients that implement OAuth 2.1 with Dynamic Client Registration need no pre-registration; everything else uses the static Authorization header.
upload_msds_pdf, create_audit_session, get_audit_report) associate activity with your account for audit traceability.Industry-sourced, AI-verified, and cryptographically signed.
| Symptom | Likely cause | Fix |
|---|---|---|
401 Unauthorized on connect | Missing or invalid API key | Check the Authorization: Bearer sk-msds-... header. Generate a fresh key at msdschain.lagentbot.com > API Keys. |
| Tools don't appear in the client | Server not loaded | Fully restart the client after adding the server, then confirm msds-chain shows in the MCP/tools list. |
| Connection drops or times out | Wrong transport | Prefer https://mcp.lagentbot.com/mcp; SSE sessions can drop when the server redeploys. GET https://mcp.lagentbot.com/health should return {"status":"ok"}. |
Credit balance exhausted | Credit allowance used up | The error says how to continue: an upgrade link and, on the free plan, when credits reset. |
| Empty results for a known chemical | Name not matched | Retry with the CAS number or a common synonym; search_chemical_database accepts name, synonym, or CAS. |
| OAuth sign-in fails | OAuth metadata unreachable | Confirm GET https://mcp.lagentbot.com/.well-known/oauth-authorization-server returns 200; the client should auto-discover the authorize and token endpoints. |
Still stuck? Email contact@lagentbot.com or open an issue at github.com/littleblakew/msds-chain-mcp/issues.
MIT
Built by LAgentBot, AI-powered chemical safety infrastructure.
Part of the MSDS Chain platform: the world's first AI Agent-driven chemical safety data trust network, powered by ChainSDS.