PH Schools MCP Server
Enables querying and analyzing the Philippine schools masterlist dataset, including searching schools, listing regions and divisions, and getting school details by BEIS ID.
README
PH Schools MCP Server
Local stdio MCP server for querying and analyzing the Philippine schools masterlist dataset.
What This Server Provides
Tools:
search_schoolsget_school_by_beis_idlist_regionslist_divisionsdataset_stats
Install and Run
npm install
npm start
Use npm start only when running from this repo manually.
Which Setup to Use
- VS Code MCP (
.vscode/mcp.json): enough for normal usage. If status isRunning, VS Code already started the server. - Claude Desktop (
claude_desktop_config.json): enough for normal usage. Restart Claude after config changes. npx -y @darwinphi/ph-schools-mcp-server ...: one-off CLI usage without cloning repo.npm install && npm start: local development/maintenance in this repository.
Usage Scenarios
mcp.jsonconfigured, no manualnpx: works for chat tool calls (dataset_stats,search_schools, etc.).- Manual
npx -y @darwinphi/ph-schools-mcp-server, no MCP client config: server process starts, but chat clients won't use it automatically. mcp.jsonconfigured plus manualnpxstart: usually unnecessary; let the MCP client manage start/stop.- One-off commands without MCP chat: use
npx ... --helpornpx ... sync-data ....
Without MCP client config, automatic VS Code/Claude tool-calling will not work.
When to Use mcp.json
Use mcp.json for normal day-to-day MCP usage in VS Code (or equivalent client config in Claude Desktop).
Use it for:
- Automatic server startup and lifecycle management by the MCP client
- MCP tool usage directly from chat prompts (
dataset_stats,search_schools, etc.) - Team/project-level shared MCP setup in a workspace
If MCP status shows Running, the client already started the server; manual npm start or manual npx start is usually unnecessary.
When to Use npx
Use npx from a terminal when you need one-off CLI actions without cloning or developing this repo.
Use it for:
- Sanity check that the published package runs:
npx -y @darwinphi/ph-schools-mcp-server --help - Manual dataset download/update:
npx -y @darwinphi/ph-schools-mcp-server sync-data --tag v1.0.1 --output "$HOME/.ph-schools/data.json" - Manual debug startup outside client-managed MCP lifecycle:
npx -y @darwinphi/ph-schools-mcp-server
Do not use npx start as a replacement for VS Code/Claude MCP config. In normal usage, let the MCP client manage server startup from its config.
When to Use npm install and npm start
Use these when working from this repository (developer/maintainer workflow), not for normal client usage.
Use them for:
- Local development while editing source files in this repo
- Running tests before commits/releases
- Debugging local unpublished changes
Typical local workflow:
npm install
npm test
npm start
If your VS Code/Claude MCP config is already working, you usually do not need to run npm start manually.
CLI (published package)
# Start MCP server over stdio
npx -y @darwinphi/ph-schools-mcp-server
# Sync canonical dataset once to a chosen path
npx -y @darwinphi/ph-schools-mcp-server sync-data --tag v1.0.1 --output "$HOME/.ph-schools/data.json"
Quick Verify
npx -y @darwinphi/ph-schools-mcp-server --help
npx -y @darwinphi/ph-schools-mcp-server sync-data --tag v1.0.1 --output "$HOME/.ph-schools/data.json"
Dataset Configuration
This server is hybrid by default:
- If local dataset file exists, it uses that file immediately.
- If local dataset file is missing, it auto-downloads from the canonical dataset URL and caches locally.
Default canonical URL (pinned tag v1.0.1):
https://raw.githubusercontent.com/darwinphi/ph-schools-dataset/v1.0.1/schools_masterlist_2020_2021.json
Runtime env vars:
PH_SCHOOLS_DATA_PATH: preferred local JSON file path (used directly if present; auto-synced to this path if missing)PH_SCHOOLS_DATA_URL: override download URL forsync-dataPH_SCHOOLS_DATA_TAG: canonical tag forsync-datawhen URL is not provided
VS Code MCP Config (Copy/Paste)
Set .vscode/mcp.json:
{
"servers": {
"phSchools": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@darwinphi/ph-schools-mcp-server"],
"env": {
"PH_SCHOOLS_DATA_PATH": "/Users/your-user/.ph-schools/data.json"
}
}
}
}
If PH_SCHOOLS_DATA_PATH file is missing, the server automatically downloads from canonical source and writes to that path.
Claude Desktop Config (Copy/Paste)
Update Claude config:
{
"mcpServers": {
"ph-schools": {
"command": "npx",
"args": ["-y", "@darwinphi/ph-schools-mcp-server"],
"env": {
"PH_SCHOOLS_DATA_PATH": "/Users/your-user/.ph-schools/data.json"
}
}
}
}
Alternative if npx is unreliable in your shell: install globally and use "command": "ph-schools-mcp-server".
If PH_SCHOOLS_DATA_PATH file is missing, the server automatically downloads from canonical source and writes to that path.
Typical macOS config file:
~/Library/Application Support/Claude/claude_desktop_config.json
Test Commands
npm test
npm run test:package
Example Prompts
Run dataset_stats and summarize key insights.List all divisions in Region I using list_divisions.Search schools in region "Region I" and division "Ilocos Norte".Get school by BEIS ID 100001 using get_school_by_beis_id.Search schools with query "High School".
Publishing and Release Flow
One-time npm setup (Trusted Publishing):
- On npmjs.com, open package
@darwinphi/ph-schools-mcp-server→ Settings → Trusted publishers. - Add GitHub Actions trusted publisher with:
- Owner/User:
darwinphi - Repository:
ph-schools-mcp-server - Workflow filename:
cd.yml
- Owner/User:
- Do not use
NPM_TOKEN; release workflow uses OIDC (id-token: write).
Manual release flow (v1):
- Update pinned dataset tag in
src/constants.js. - Bump package version:
npm version patch # or minor / major
What npm version patch does:
- Updates
package.jsonversion (e.g.,1.0.1->1.0.2) - Updates
package-lock.jsonversion fields - Creates a git commit
- Creates a git tag (e.g.,
v1.0.2)
- Sync
server.jsonversion to matchpackage.json. - Run:
npm ci
npm test
npm run test:package
- Push commit and tags:
git push
git push --tags
- Create GitHub Release notes from the tag:
gh release create v<new_version> --generate-notes --title "v<new_version>"
Example:
gh release create v1.0.2 --generate-notes --title "v1.0.2"
- Push commit and tags. Tag pushes (
v*) automatically triggerCD Releaseworkflow. - Workflow publishes npm package, then publishes MCP Registry metadata.
For metadata-only updates, use workflow Publish MCP Registry Metadata.
License and Data Provenance
- Code license: ISC (LICENSE).
- Dataset source:
darwinphi/ph-schools-dataset(canonical repository backed by gov.ph source data). - Use of dataset remains subject to source terms and applicable policies.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.