ESG Hub MCP Server
RemoteAbout ESG Hub MCP Server
ESG Hub MCP Server gives AI agents access to an ESG knowledge base spanning standards, regulations, frameworks and regional guidance. Its MCP tools search and read curated articles and resources.
Key features
- 01Search ESG articles and resources
- 02Read knowledge base content
- 03Browse frameworks, industries and terms
- 04Access ESG guidance across regions
Use cases
- 01Find guidance on ESG standards
- 02Look up regional ESG requirements
- 03Explore frameworks and industry topics
MCP support
The MCP server exposes tools for AI agents to search and read the ESG Hub knowledge base.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://esg-hub.ascent.partners/api/mcpclaude mcp add --transport http com-ascent-partners-esg-hub https://esg-hub.ascent.partners/api/mcp{"mcpServers":{"com-ascent-partners-esg-hub":{"type":"http","url":"https://esg-hub.ascent.partners/api/mcp"}}}Open in Cursor{"servers":{"com-ascent-partners-esg-hub":{"type":"http","url":"https://esg-hub.ascent.partners/api/mcp"}}}Open in VS Code{"mcpServers":{"com-ascent-partners-esg-hub":{"type":"http","url":"https://esg-hub.ascent.partners/api/mcp"}}}{"mcpServers":{"com-ascent-partners-esg-hub":{"serverUrl":"https://esg-hub.ascent.partners/api/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
15This tool takes no parameters.
Details
From the project READMEESG Hub
An open-access encyclopedia of Environmental, Social, and Governance (ESG) knowledge by Ascent Partners Foundation — 350+ articles, 240+ curated resources, books, courses, AI search, a public REST API, and an MCP server. Available in English, Simplified Chinese, and Traditional Chinese (Hong Kong).
Live site: esg-hub.ascent.partners
Overview
- Knowledge base — 350+ ESG articles across standards (GRI, IFRS S1/S2, TCFD, TNFD, ESRS), regulations, frameworks, ratings, and regional (HK/APAC) guidance
- AI search — ask questions in natural language; answers grounded in the knowledge base (SurrealDB BM25 + vector search, DeepSeek, Brave Search across authoritative ESG domains)
- Public REST API — open access to all content:
/api/v1 - MCP server — connect AI agents directly:
@esg-hub/mcp-server - Learning resources — free courses, books, and a curated video library
- i18n — English, Simplified Chinese (zh), Traditional Chinese — Hong Kong (hk)
Architecture
flowchart TB
U["User / AI agent"] --> WEB["Next.js 16 App Router · React 19<br/>next-intl (en / zh / hk)"]
WEB --> DB[("SurrealDB Cloud<br/>BM25 + vector search")]
WEB --> AI["DeepSeek + Brave Search<br/>grounded answers"]
WEB --> API["Public REST API · /api/v1"]
API --> MCP["MCP server<br/>@simonmak-ascent/esg-hub-mcp"]
V["APG ALB · esg-hub.ascent-partners.com"] --> WEB
How a query is answered
sequenceDiagram
autonumber
actor U as User
participant W as Next.js app (App Router)
participant API as /api/ai-search · /api/ai-chat
participant S as search.ts
participant DB as SurrealDB (page · section · embedding)
participant AI as DeepSeek
participant X as Brave / OpenAlex
U->>W: ask a question
W->>API: POST query (+ locale)
API->>S: search(query)
S->>DB: BM25 keyword + vector (embedding) hybrid ranking
DB-->>S: ranked pages / sections + citations
API->>AI: synthesize answer from ranked context
API->>X: enrich (web + research)
X-->>API: references
API-->>W: streamed answer + citations
W-->>U: grounded response
Data model
erDiagram
PAGE ||--o{ SECTION : contains
SECTION ||--o{ EMBEDDING : "vectorised as"
FRAMEWORK ||--o{ PAGE : classifies
INDUSTRY ||--o{ PAGE : tags
TERM ||--o{ PAGE : defines
PAGE {
string id
string slug
string title
string lang
}
SECTION {
string id
string heading
text body
}
EMBEDDING {
vector vector
string model
}
FRAMEWORK, INDUSTRY and TERM are the facets exposed by the /api/v1/* REST
resources; page, section and embedding are the core SurrealDB tables.
Deployment
Production runs on the ecs-01/02/03 HA fleet as the digest-pinned container declared in
container.json — the image is the deployable, not a release tree on a node. The fleet runs it as
a container on :3210, and no node runs a routing web server. The ALB terminates TLS and routes
esg-hub.ascent.partners to the container's port. The former Vercel project was retired. See
apg-infra for the fleet deployment.
flowchart LR
DEV["git push / PR"] --> CI["GitHub Actions<br/>lint · test · e2e · health-check"]
CI -->|merge to main| IMG["digest-pinned image<br/>container.json"]
IMG -->|the image is the deployable| PROD["container · :3210"]
ALB["APG ALB · TLS termination"] -->|esg-hub.ascent.partners| PROD
PROD --> SURF["/api/mcp · /api/v1<br/>MCP + REST surfaces"]
Tech Stack
Next.js 16 (App Router, React 19) · SurrealDB Cloud · Tailwind CSS v4 · next-intl · Vitest + Playwright · APG fleet (containerised)
Quick Start (≤ 5 minutes)
Fastest path — nothing to install: use the live site and the open API.
- Site: https://esg-hub.ascent.partners
- API (no auth):
GET https://esg-hub.ascent.partners/api/v1/pages
Run a local copy:
git clone https://github.com/simonmak-ascent/esg-hub.git
cd esg-hub
pnpm install # note: requires sibling repo ../tool_package (private)
cp .env.example .env # fill in your own values
pnpm dev
Useful commands: pnpm lint · npx tsc --noEmit · npx vitest run · pnpm verify:db
Note: the app renders content from a SurrealDB database — you need your own instance and content to run a full copy. The public site and API are freely accessible without running anything.
API & MCP
- REST API: open, no auth —
GET /api/v1/pages,/api/v1/search?q=...,/api/v1/resources,/api/v1/meta(docs) - MCP server: see
mcp-server/— gives AI agents tools to search and read the knowledge base
Contributing
- Content errors (wrong/outdated info, dead links): open a content error issue
- Bugs and code changes: see CONTRIBUTING.md
- Security vulnerabilities: report privately via SECURITY.md
Use with Context7
Up-to-date ESG Hub documentation is indexed on Context7, so coding agents can pull it into context on demand. With the Context7 MCP server or ctx7 CLI installed, name the library in your prompt:
use library /simonmak-ascent/esg-hub for API and docs
License
- Code: MIT
- Encyclopedia content: CC BY-SA 4.0 by Ascent Partners Foundation
An Ascent Partners Foundation project.
If this saves you time, a ⭐ on GitHub helps others find it.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.