This server provides AI assistants with comprehensive international education data from UNESCO UIS (4,000+ indicators, all member countries) and OECD Education at a Glance (38 countries + partners), requiring no API keys.
UNESCO UIS Capabilities
Search indicators (uis_list_indicators): Browse/filter 4,000+ indicators by theme (literacy, enrollment, expenditure, teacher ratios, etc.) or keyword
List countries & regions (uis_list_countries): Retrieve ISO 3166-1 Alpha-3 codes and regional aggregates (World Bank income groups, SDG regions)
Fetch education data (uis_get_education_data): Pull time-series data for any indicator for one or all countries, with optional year range filtering
Compare countries (uis_compare_countries): Side-by-side comparison of up to 15 countries on a single indicator, sorted by value
Country education profile (uis_country_education_profile): Full profile covering 10 core indicators (literacy, enrollment, completion, expenditure, teacher ratios, gender parity) in one call
List database versions (uis_list_versions): Check available UNESCO UIS release versions and publication dates
OECD Capabilities
List OECD datasets (oecd_list_education_datasets): Browse Education at a Glance dataflows (spending, enrollment, graduation, teacher salaries, labour market outcomes)
Fetch OECD indicator data (oecd_get_education_indicator): Retrieve SDMX data for specific dataflows filtered by country and time period
Search OECD datasets (oecd_search_datasets): Keyword search across all OECD SDMX dataflows
Cross-Source & Additional Features
Benchmark countries (education_benchmark_countries): Multi-country benchmark across five themes — literacy, spending, completion, teachers, or enrollment
SDG-4 monitoring prompt: Built-in structured reporting for Education for All goals
Switzerland comparison prompt: Pre-built comparison of Switzerland vs. Finland, Singapore, and Japan
Reference resources: Quick-reference guides for core UNESCO indicator IDs and OECD dataflow IDs
No API keys required: Both data sources are open and publicly accessible
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@global-education-mcp Compare education spending as % of GDP for Switzerland and Finland"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
🇨🇭 Part of the Swiss Public Data MCP Portfolio
MCP server for international education data – UNESCO UIS (4,000+ indicators across all member countries) and OECD Education at a Glance via SDMX. No API keys required.
global-education-mcp gives AI assistants like Claude a complete international education intelligence system – literacy rates, enrolment ratios, education expenditure, teacher salaries, gender parity and SDG-4 monitoring, all accessible through a single standardised MCP interface.
The server bridges two of the most authoritative sources for internationally comparable education statistics: UNESCO UIS (global coverage, 4,000+ indicators) and the OECD's annual Education at a Glance (38 OECD countries, SDMX REST API). Both are open and require no API key.
Anchor demo query: "Compare Switzerland's education expenditure as a percentage of GDP with Finland, Singapore and South Korea over the last 10 years – and flag any SDG-4 gaps."
Related MCP server: Skolverket-MCP
🌍 UNESCO UIS – 4,000+ indicators, all UNESCO member countries, no API key
📊 OECD Education at a Glance – 38 OECD countries + partners via SDMX REST
🔍 Indicator search – browse and filter the full UNESCO indicator catalogue
🗺️ Multi-country comparison – benchmark any indicator across multiple countries
🏫 Country education profiles – 10 core indicators in one call
🎯 SDG-4 monitoring – structured reporting on Education for All targets
📈 OECD dataset search – discover and retrieve Education at a Glance dataflows
🔑 No API keys required – fully open data, zero setup friction
☁️ Dual transport – stdio for Claude Desktop, Streamable HTTP/SSE for cloud deployment
🛡️ Graceful degradation – API failures return helpful messages with local reference fallback
Python 3.11+
uv (recommended) or pip
No API keys needed
# Clone the repository
git clone https://github.com/malkreide/global-education-mcp.git
cd global-education-mcp
# Install
pip install -e ".[dev]"Or with uvx (no permanent installation):
uvx global-education-mcp# Start the server (stdio mode for Claude Desktop)
global-education-mcpTry it immediately in Claude Desktop:
"What is Switzerland's literacy rate compared to Finland and Singapore?" "Show me education expenditure as % of GDP for CHE, DEU and AUT over the last 10 years."
Windows (%APPDATA%\Claude\claude_desktop_config.json):
{
"mcpServers": {
"global-education": {
"command": "uvx",
"args": ["global-education-mcp"]
}
}
}macOS (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"global-education": {
"command": "uvx",
"args": ["global-education-mcp"]
}
}
}A ready-to-use claude_desktop_config.json is included in the repository root.
For use via claude.ai in the browser (e.g. on managed workstations without local software).
⚠️ Security note: Since v0.3,
MCP_HOSTdefaults to127.0.0.1. The SSE transport must always run behind a reverse proxy that adds TLS, authentication, and rate-limiting. Never expose the raw port to the internet —MCP_HOST=0.0.0.0is only safe inside an isolated container network.
Docker (recommended):
The repository ships a hardened multi-stage Dockerfile and a
docker-compose.yml that applies read_only: true, cap_drop: [ALL],
security_opt: [no-new-privileges:true], and runs as non-root user
uid 10001. The compose file binds the port to 127.0.0.1 so a host-level
reverse proxy is required for any external access.
docker compose up --build
# then point nginx/caddy at 127.0.0.1:8000/sse with TLS + authPlain docker run (without compose):
docker build -t global-education-mcp .
docker run --rm \
--read-only --cap-drop ALL --security-opt no-new-privileges \
--tmpfs /tmp:size=16M,mode=1777 \
-p 127.0.0.1:8000:8000 \
global-education-mcpRender.com:
Push/fork the repository to GitHub
On render.com: New Web Service → connect GitHub repo
Set environment variables in the Render dashboard:
MCP_TRANSPORT=sse
MCP_HOST=0.0.0.0 # Render needs 0.0.0.0; their edge layer provides TLS + auth.
PORT=8000In claude.ai under Settings → MCP Servers, add: https://your-app.onrender.com/sse
💡 "stdio for the developer laptop, sandboxed SSE container for the browser."
Tool | Description |
| Search and list available indicators (4,000+) |
| List countries and regions with ISO codes |
| Retrieve data for a specific indicator |
| Multi-country comparison for one indicator |
| Full education profile (10 core indicators) |
| List available database versions |
Tool | Description |
| List Education at a Glance datasets |
| Retrieve OECD education data via SDMX |
| Search OECD dataflows by keyword |
Tool | Description |
| Benchmark multiple countries across 5 focus themes (UNESCO UIS) |
Resources:
education://indicators/unesco – Quick reference for core UNESCO indicators
education://datasets/oecd – Quick reference for OECD Education at a Glance dataflows
Prompts:
bildungsvergleich_schweiz – Switzerland vs. Finland, Singapore, Japan
sdg4_monitoring – SDG-4 report for CH/DE/AT
ISO 3166-1 Alpha-3 standard:
Code | Country | Code | Country |
| Switzerland |
| Finland |
| Germany |
| Singapore |
| Austria |
| South Korea |
| France |
| Japan |
| Sweden |
| United States |
Query | Tool |
"What is Switzerland's literacy rate vs. Finland and Singapore?" |
|
"Education expenditure as % of GDP for CHE, DEU, AUT over 10 years" |
|
"Create a full education profile for South Korea" |
|
"Which OECD datasets cover teacher salaries?" |
|
"Compare secondary graduation rates across 5 European countries" |
|
"Create an SDG-4 monitoring report for Switzerland" |
|
→ More use cases by audience →
┌─────────────────┐ ┌──────────────────────────────┐ ┌────────────────────┐
│ Claude / AI │────▶│ Global Education MCP │────▶│ UNESCO UIS API │
│ (MCP Host) │◀────│ (MCP Server) │◀────│ uis.unesco.org │
└─────────────────┘ │ │ └────────────────────┘
│ 10 Tools · 2 Resources │
│ · 2 Prompts │ ┌────────────────────┐
│ Stdio | SSE │────▶│ OECD SDMX API │
│ │◀────│ sdmx.oecd.org │
│ server.py │ └────────────────────┘
│ + api_client.py │
└──────────────────────────────┘Component | Metaphor | Function |
HTTPClient | Postal service | Handles all outbound HTTP requests, retries and timeouts |
SimpleCache | Whiteboard | In-memory TTL cache for repeated queries |
GracefulFallback | Safety net | Returns local reference data when APIs are unavailable |
SDMXParser | Translator | Converts OECD SDMX/XML responses to clean JSON |
Data Source | Cache TTL | Rationale |
UNESCO UIS indicators | 3600s | Catalogue is stable; updated annually |
UNESCO UIS country data | 1800s | Figures update yearly, not intraday |
OECD dataset list | 3600s | Education at a Glance is an annual publication |
OECD indicator data | 1800s | Same annual update cycle |
Country/region list | 86400s | ISO codes and country lists are highly stable |
global-education-mcp/
├── src/global_education_mcp/ # Main package
│ ├── __init__.py # Package metadata, version
│ ├── server.py # FastMCP server, 10 tools, 2 resources, 2 prompts
│ └── api_client.py # HTTP client, UNESCO UIS + OECD wrappers, formatters
├── scripts/
│ └── record_fixtures.py # Records the fixtures from the live UIS API
├── tests/
│ ├── fixtures/ # Recorded responses + PROVENANCE.md (source, date, SHA-256)
│ ├── fixture_data.py # Fixture loader – raises on a missing name
│ ├── test_source_contract.py # Contract vs. the recording + live tests
│ ├── test_server.py # 42 tests (basic / intermediate / advanced)
│ └── test_extended_scenarios.py # 74 tests across 8 categories
├── claude_desktop_config.json # Ready-to-use Claude Desktop config
├── pyproject.toml # Build configuration (hatchling)
├── CHANGELOG.md
├── CONTRIBUTING.md # Contribution guide (English)
├── CONTRIBUTING.de.md # Contribution guide (German)
├── SECURITY.md # Security policy (English)
├── SECURITY.de.md # Security policy (German)
├── LICENSE
├── README.md # This file (English)
└── README.de.md # German versionUNESCO UIS: Some indicators have sparse coverage for low-income countries or recent years
OECD SDMX: Occasional API timeouts on large multi-country, multi-year requests; reduce the year range if needed
OECD coverage: 38 OECD members + select partners – does not cover all UNESCO member states
Historical depth: UNESCO UIS data availability varies by indicator; not all series go back to 1970
Language: UNESCO UIS returns indicator labels in English only; OECD labels may vary by dataflow
No real-time data: Both sources publish annually – figures reflect the latest published edition, not live school statistics
Estimated values are labelled: UIS marks part of its observations UIS_EST (UIS estimate) or NAT_EST (national estimate) – for LR.AG15T99 that is 1,306 of 9,818 values. The status column names it; an unlabelled value is a reported one.
Data rows carry the ISO code, not the country name: plain names come from uis_list_countries. The data tools print the code rather than inventing a name.
An unknown country code is not an error status: UIS answers HTTP 200 with an empty result set and states the reason in the payload's hints field. The tools print that hint – otherwise a typo would look exactly like a country without data.
Verbindliche Klassifikation für den Einsatz im Schulamt der Stadt Zürich (German section follows in
README.de.md).
Dimension | Class | Reasoning |
Confidentiality | public | UNESCO UIS data licensed under CC BY-SA 3.0 IGO; OECD EaG under public OECD Terms |
Integrity | normal | Upstream is authoritative; local in-memory cache is TTL-bounded and never written to disk |
Availability | normal | Graceful fallback to bundled static reference data when an API is unreachable |
Overall protection class | G1 — public (öffentlich) | lowest tier per ISDS Stadt Zürich |
Data owner: UNESCO UIS / OECD (external)
System owner: Schulamt der Stadt Zürich
Processes personal data: no
DSG / EDÖB relevance: none (only anonymized country-level aggregates)
Aspect | Classification |
Tool output (Markdown tables, summaries) | BUI (betrieblich unkritische Information) |
In-memory TTL cache | BUI (same tier as source) |
Structured logs (JSON on stderr, see OBS-003) | BUI — only tool name, params, duration; no PII |
| BUI |
→ The server is approved for any Schulamt use case without additional clearance from the data protection officer.
Component | Supported version |
MCP Protocol | 2024-11-05 |
MCP Python SDK |
|
Python | 3.11, 3.12, 3.13 |
|
|
|
|
Major-version upgrades are deliberate decisions — the upper bounds in
pyproject.toml exist so a transitive bump does not silently break the
server. See CHANGELOG.md for the upgrade trail.
Aspect | Details |
Access | Read-only ( |
Personal data | No personal data — UNESCO UIS and OECD EaG publish only aggregated, country-level statistics |
Rate limits | Built-in per-query caps (max 50 indicators per search, max 10 countries per comparison, conservative year ranges) |
Caching | In-memory TTL cache (1800–86400s) reduces upstream load and respects publisher capacity |
Timeout | 30 seconds per upstream API call, with graceful fallback to local reference data |
Authentication | No API keys required — both UNESCO UIS and OECD SDMX are publicly accessible |
Licenses | UNESCO UIS data under CC BY-SA 3.0 IGO; OECD data under OECD Terms and Conditions |
Terms of Service | Subject to ToS of the respective sources: UNESCO UIS, OECD — please cite the source when redistributing |
Attribution | All tool responses include source attribution ( |
This server speaks two protocol eras over the same endpoint. The client's first request on a connection decides which one applies; a later claim from the other era is refused.
Era | Revision | Who reaches it |
|
| What today's clients speak. The server answers with the revision asked for, or with the |
Per-request envelope |
| A request carrying the |
Both revisions are pinned in
tests/test_protocol_version.py and asserted
against the installed SDK, so a Dependabot bump of mcp cannot move either one
silently. This server builds no ASGI app to send an initialize through, so
the gate asserts the SDK constants rather than a measured response — the
weaker form, named rather than left unsaid.
Note that the SDK's LATEST_PROTOCOL_VERSION is an alias for the modern
era, not for the handshake era — pinning against it alone would leave the era
that current clients actually negotiate free to drift.
Update policy. When the gate fails, do not edit the constant blindly: read
the spec changelog between the two revisions, verify the server still behaves,
then move the constant, this section, README.de.md and
CHANGELOG.md together.
# Unit tests (no API key required, no network)
PYTHONPATH=src pytest tests/ -v -m "not integration"
# Full suite including the live tests against api.uis.unesco.org
PYTHONPATH=src pytest tests/ -v
# Re-record the fixtures (writes tests/fixtures/ + PROVENANCE.md)
PYTHONPATH=src python scripts/record_fixtures.py169 tests – 152 offline, 17 against the live source.
Category | Tests | Description |
Edge cases & boundary values | 19 | Year limits, string lengths, null/zero values |
Security & adversarial inputs | 14 | Injection attempts, HTTP error codes, whitespace |
Output quality | 11 | Markdown structure, source attribution, sort order |
Resilience & error cascades | 9 | Full API outage, partial results, timeouts |
Subject-matter correctness | 10 | SDG-4 coverage, correct indicators per focus theme |
Performance & concurrency | 4 | Concurrent requests, time limits |
Schulamt scenarios | 7 | DACH comparison, PISA, teacher shortage |
Source contract vs. the recording | 23 | Field names, envelope, query parameters, hints |
Live tests ( | 17 | The paths, the shape, and the tools themselves |
A hand-written mock encodes its author's assumption and therefore cannot refute it: production code and fixture come from the same head, the same hour, the same reading of the docs. Where both are wrong, both are wrong together — and the suite stays green.
That is not a hypothetical here. Before 2026-08-08 this repo had 128 green
tests while three of its four UNESCO paths answered HTTP 404 and every data
query returned an empty list. The mocks carried the same invented field names
as the production code (observations, indicatorId, entityType), so
nothing could ever contradict them.
Every fixture under tests/fixtures/ is now a recorded response.
PROVENANCE.md names the source URL, the recording date, the selection rule
and the SHA-256 for each one. Without a date, "recorded" becomes
indistinguishable from "invented" after two years — the file looks the same.
Three of the fixtures are controls: a made-up theme value, a made-up
country code, and the same time series requested twice with different
parameter names. Without them a measurement only shows what we received; with
them it shows what the source actually distinguishes.
Contributions are welcome. Please open an issue first to discuss what you would like to change.
Follow the existing code style (Ruff linting, Black formatting)
Add tests for new tools (tests/test_server.py or test_extended_scenarios.py)
Use the @pytest.mark.integration marker for tests that call live APIs
Update CHANGELOG.md and the tool table in this README
See CONTRIBUTING.md for the full contribution guide
See CHANGELOG.md
To report a vulnerability, see SECURITY.md (🇩🇪 Deutsche Version). Please use the private channels described there rather than public issues.
MIT License — see LICENSE
Hayal Oezkan · github.com/malkreide
Data: UNESCO Institute for Statistics (UIS) – open education data for all UNESCO member states
Data: OECD Education at a Glance – annual OECD education statistics via SDMX
Protocol: Model Context Protocol – Anthropic / Linux Foundation
Related: swiss-transport-mcp – MCP server for Swiss public transport
Related: zurich-opendata-mcp – MCP server for Zurich city open data
Portfolio: Swiss Public Data MCP Portfolio
Run via uv's uvx — no clone or manual install needed. Add to your MCP client config (mcpServers for Claude Desktop, Cursor and Windsurf; use a top-level servers key for VS Code in .vscode/mcp.json):
{
"mcpServers": {
"global-education-mcp": {
"command": "uvx",
"args": [
"global-education-mcp"
]
}
}
}Benchmarkt mehrere Länder auf einem Bildungsthema via UNESCO UIS.
Automatisch werden die passenden Indikatoren für den gewählten Fokus ausgewählt und ein strukturierter Vergleich erstellt.
Fokus-Optionen:
literacy: Alphabetisierungsraten (Erwachsene + Jugendliche)
spending: Bildungsausgaben (% BIP, verschiedene Stufen)
completion: Abschlussquoten (Primar, Sek I, Sek II)
teachers: Schüler-Lehrer-Verhältnis + Lehrerausbildung
enrollment: Einschulungsraten nach Schulstufe
Args: params: country_codes, focus
Returns: Vollständiger Benchmarkreport mit Tabellen für alle relevanten Indikatoren
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already mark the tool as read-only and non-destructive. The description adds behavioral context: it automatically selects appropriate indicators based on focus and generates a structured report. This goes beyond annotations but does not detail other potential side effects.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and well-structured, using bullet points for focus options. It front-loads the core purpose and then details parameters and output efficiently. Every sentence adds value.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the presence of an output schema and annotations, the description covers purpose, parameters, and output format. It lacks discussion of error handling or prerequisites but is adequate for a non-destructive benchmarking tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Despite schema description coverage reported as 0%, the description compensates by explaining the focus parameter with detailed options and linking to the required country_codes parameter. It adds meaning beyond the raw schema by providing context for each focus area.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool benchmarks multiple countries on a specific education topic using UNESCO UIS. It lists five distinct focus areas and contrasts with siblings by emphasizing automated indicator selection for structured comparison.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for benchmarking on specific topics but does not explicitly state when to use this tool versus alternatives like uis_compare_countries. No 'when-not-to-use' guidance or references to sibling tools are provided.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Ruft Bildungsdaten aus dem OECD Education at a Glance Report ab.
Greift auf die OECD SDMX REST API zu. Liefert strukturierte Daten für OECD-Länder zu Bildungsausgaben, Einschreibungsraten, Lehrergehältern etc.
Beispiele:
Bildungsausgaben Schweiz/DE/AT: dataflow='EAG_FISC', countries=['CHE','DEU','AUT']
Lehrergehälter OECD: dataflow='EAG_PERS_SALARY'
Beschäftigung nach Bildungsabschluss: dataflow='EAG_EMP_EDUC'
Args: params: dataflow_id, countries (optional), start_period, end_period
Returns: Markdown-formatierte Datentabelle oder Rohdaten-Zusammenfassung
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and destructiveHint=false, so the tool is clearly non-destructive. The description adds context about accessing the OECD API and returning structured data, but does not go beyond what annotations imply. No contradictions are present.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is reasonably concise and well-structured, with a clear introduction, helpful examples, and a brief listing of arguments. It front-loads the main purpose and uses bullet points for examples. It could be slightly shorter without losing key information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity and the existence of an output schema (though not provided here), the description adequately covers the return format as markdown table or summary. It also mentions available dataflow IDs. However, it omits potential error conditions or dataset size warnings.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema provides detailed descriptions for each parameter, so the description adds limited value by simply listing parameter names. The description mentions 'dataflow_id' and 'countries' but does not elaborate on their meaning beyond what the schema already provides. With high schema coverage, a baseline of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states that the tool retrieves education data from OECD Education at a Glance report via SDMX REST API. It provides specific examples of dataflows and countries, making the purpose evident. However, it does not explicitly differentiate from sibling tools like uis_get_education_data, which serve similar but distinct data sources.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description offers examples of usage but lacks explicit guidance on when to use this tool versus alternatives. It does not mention prerequisites, exclusions, or scenarios where sibling tools might be more appropriate. This omission reduces clarity for an AI agent deciding which tool to invoke.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Listet verfügbare OECD Education at a Glance Datensätze auf.
Education at a Glance ist das jährliche Referenzwerk der OECD für internationale Bildungsvergleiche. Es umfasst 38 OECD-Länder plus Partner.
Abgedeckte Themenbereiche:
Bildungsbeteiligung und -abschlüsse (EAG_ENRL, EAG_GRAD_ENTR)
Bildungsausgaben und Finanzierung (EAG_FISC)
Lehrpersonal und Arbeitsbedingungen (EAG_PERS, EAG_PERS_SALARY)
Bildungsrendite und Arbeitsmarkt (EAG_EMP_EDUC, EAG_EARN_RATIO)
Returns: Markdown-Liste der Datensätze mit Beschreibungen und Dataflow-IDs
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already convey read-only, idempotent, and open-world nature. The description adds value by specifying the return format (Markdown list with descriptions and dataflow IDs) and the scope (OECD countries plus partners), which goes beyond the annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured with bullet points and clear sections. It front-loads the main purpose. Slightly verbose in the topic area listing but overall concise and easy to parse.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given no parameters, an output schema exists, and rich annotations, the description covers purpose and return value well. However, it lacks guidance on when to use relative to siblings, which slightly detracts from completeness.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, and the schema coverage is 100% (trivially). The description does not need to add parameter semantics, and the baseline for 0 parameters is 4. It doesn't detract from this score.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it lists available OECD Education at a Glance datasets, specifying the resource and action. It enumerates topic areas with example dataflow IDs, effectively distinguishing it from sibling tools like oecd_get_education_indicator or uis_list_countries.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description does not provide explicit guidance on when to use this tool versus alternatives. No context about prerequisites or exclusions, such as when a more specific tool like oecd_search_datasets might be appropriate.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Durchsucht alle OECD-Datensätze nach einem Stichwort.
Über die OECD SDMX API sind hunderte Datensätze verfügbar. Diese Funktion findet Datensätze mit Bildungsbezug oder anderen Themen.
Args: params: keyword, limit
Returns: Markdown-Liste gefundener Datensätze mit Dataflow-IDs
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, destructiveHint=false, idempotentHint=true, openWorldHint=true. The description adds value by noting the API source ('OECD SDMX API'), that it searches hundreds of datasets, and that the output is a Markdown list with Dataflow-IDs. No contradictions with annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured with a lead sentence stating the purpose, followed by context, then an Args/Returns section. It is concise and contains no superfluous information, earning a high score.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity (search with two parameters, output as list), the description covers purpose, data source, and output format. An output schema exists, so explaining return values is unnecessary. Missing details like error handling or empty results are minor gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description barely adds value beyond the input schema: it only lists parameter names ('keyword, limit') without describing their purpose or constraints. The schema itself already contains good descriptions for each parameter, so the description's contribution is minimal, leading to a low score.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Durchsucht alle OECD-Datensätze nach einem Stichwort', identifying the verb (search) and resource (all OECD datasets) with a specific action (by keyword). This distinguishes it from siblings like oecd_list_education_datasets, which likely lists all without searching.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides context about the SDMX API and that it finds datasets related to education or other topics, but it does not explicitly state when to use this tool versus alternatives. There is no mention of prerequisites or when not to use it, leaving differentiation to the agent's inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Vergleicht Bildungsindikatoren zwischen mehreren Ländern (UNESCO UIS).
Ideal für den direkten internationalen Vergleich: Wie steht die Schweiz im Vergleich zu Finnland, Singapur und dem OECD-Durchschnitt?
Args: params: indicator_id, country_codes (Liste), year (optional)
Returns: Markdown-Vergleichstabelle sortiert nach Indikatorwert
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description adds that it returns a Markdown comparison table sorted by indicator value, which is useful beyond the annotations (readOnlyHint, idempotentHint). It does not contradict any annotation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise (3 sentences plus args/return) and front-loaded with the main purpose. It could be more structured (e.g., bulleted info) but is efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the presence of an output schema (though not shown) and the complexity of the tool (comparison with country list), the description adequately covers input, output format, and use case. It lacks error handling details but is sufficient for selection.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description lists the parameters (indicator_id, country_codes, year) with minimal additional context (e.g., year optional, country codes as list). The schema already provides detailed descriptions for each field, so the description adds limited value.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool compares educational indicators across countries using UNESCO UIS data, and provides an example (Switzerland vs Finland etc.). It is specific and actionable, but does not explicitly differentiate from sibling tools like education_benchmark_countries.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The phrase 'Ideal für den direkten internationalen Vergleich' suggests when to use, but there is no guidance on when not to use or mention of alternative tools (e.g., uis_get_education_data for single country queries).
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Erstellt ein umfassendes Bildungsprofil für ein Land via UNESCO UIS.
Ruft automatisch die wichtigsten Schlüsselindikatoren ab: Alphabetisierung, Einschulungsraten, Abschlussquoten, Bildungsausgaben, Schüler-Lehrer-Verhältnis, Geschlechterparität.
Args: params: country_code (ISO Alpha-3), latest_year_only
Returns: Vollständiges Markdown-Bildungsprofil mit allen Kernindikatoren
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already provide readOnlyHint and idempotentHint. Description adds that it returns a Markdown profile with specific indicator types, but no additional behavioral details (e.g., no mention of side effects, auth, or rate limits).
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Concise with short bullet list of indicators, clear Args and Returns sections. Efficient use of space, no fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, the description adequately covers the tool's purpose and return format (Markdown profile). No major gaps, though edge cases like missing data could be addressed.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema descriptions already cover parameter details (country_code format, latest_year_only toggle). Description adds value by listing the types of indicators included (literacy, enrollment, etc.), which aids understanding beyond schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it creates a comprehensive education profile for a country using UNESCO UIS, listing specific indicators. It distinguishes from sibling tools like uis_get_education_data (raw data) and uis_compare_countries (comparison).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Implied usage for obtaining a full country profile, but no explicit guidance on when to use vs alternatives (e.g., uis_get_education_data for raw data). No when-not-to-use or prerequisite conditions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Ruft Bildungsdaten von der UNESCO Institute for Statistics API ab.
Kernfunktion des Servers. Liefert international vergleichbare Daten zu einem spezifischen Indikator – für ein Land oder alle Länder.
Typische Anwendungsfälle:
Alphabetisierungsrate der Schweiz: indicator='LR.AG15T99', country='CHE'
Bildungsausgaben OECD-Länder: indicator='XGDP.FSGOV', kein Land
Schulabschlussquoten Europa: indicator='CR.1' + Jahresfilter
Schüler-Lehrer-Verhältnis: indicator='PTR.1'
Args: params: indicator_id (erforderlich), country_code, start_year, end_year
Returns: Markdown-formatierte Tabelle oder Zeitreihe mit Daten und Metadaten
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, destructiveHint=false, idempotentHint=true. The description adds that the result is a markdown-formatted table or time series, and warns that omitting a country can return a large dataset. This provides useful behavioral context beyond annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured with a clear purpose, examples, and parameter list. It is moderately sized and front-loaded. Some redundancy with the schema exists, but overall it is efficient and easy to parse.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the complexity of the tool (nested parameters, many sibling tools), the description covers purpose, usage examples, parameter list, and return format. It references sibling tools indirectly in the parameter description. With annotations and output schema present, it is fairly complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Although the input schema has detailed descriptions inside $defs, the top-level schema coverage is 0%. The description lists parameters and gives examples, but does not explain each parameter in depth. It partially compensates for the low schema coverage but not fully.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves educational data from the UNESCO Institute for Statistics API, specifying it is the core function of the server and providing concrete examples with expected outputs. It distinguishes itself from siblings by focusing on raw indicator data for one or all countries.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives typical use cases with example parameters, but does not explicitly tell when to avoid this tool in favor of siblings like uis_compare_countries or uis_list_indicators. The context of usage is implied but lacks explicit exclusion criteria.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Listet verfügbare Länder und Regionen in der UNESCO UIS-Datenbank auf.
Gibt ISO 3166-1 Alpha-3 Codes zurück, die für Datenabfragen benötigt werden. Beispiele: CHE (Schweiz), DEU (Deutschland), AUT (Österreich), FRA (Frankreich).
Neben Einzelländern sind auch regionale Aggregate verfügbar: Weltregionen, Einkommensgruppen (World Bank), SDG-Regionen.
Args: params: search (optional Textfilter), entity_type (optional)
Returns: Markdown-Liste mit ISO-Codes und Ländernamen
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, destructiveHint, idempotentHint. The description adds that it returns ISO codes and provides examples, plus the Args/Returns section. No contradictions, and it adds useful context beyond annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, front-loads the main purpose, includes examples, and is well-structured with paragraphs and an Args/Returns block. No unnecessary information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the presence of an output schema and low tool complexity, the description provides sufficient context: purpose, return format, examples, and types of regions. It is complete for agent usage.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema already contains descriptions for both parameters (search and entity_type). The description briefly repeats them in the Args section but does not add significant new meaning. Schema description coverage is 0% according to signal, but actual schema descriptions exist, so the description adds minimal value.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: listing available countries and regions from UNESCO UIS database, returning ISO codes. It distinguishes from sibling tools like uis_compare_countries and uis_get_education_data by focusing on the list of regions.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains that the tool is used to obtain ISO codes for data queries and lists various region types. It does not explicitly state when not to use, but the context is clear and sufficient.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Listet verfügbare Indikatoren der UNESCO Institute for Statistics auf.
Die UIS bietet über 4'000 Indikatoren zu Bildung, Wissenschaft, Kultur und Kommunikation. Diese Funktion dient zur Exploration und Indikatorsuche.
Wichtige Indikator-Kategorien:
Alphabetisierung (LR.*): Lese-/Schreibkompetenz nach Alter, Geschlecht
Einschulungsraten (NERA.*): Netto-Einschulungsraten nach Schulstufe
Schulabschlüsse (CR.*): Abschlussquoten Primar- bis Sekundarstufe
Bildungsausgaben (XGDP., XUNIT.): % BIP, pro Schüler
Lehrerquoten (PTR., TRTP.): Schüler-Lehrer-Verhältnis, Ausbildungsgrad
Args: params: theme (optional), search (optional), limit
Returns: Markdown-Liste mit Indikator-IDs und Beschreibungen
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, destructiveHint=false, idempotentHint=true, openWorldHint=true. The description adds behavioral context by explaining it's for exploration and listing categories, but does not disclose additional traits like pagination or rate limits. No contradiction with annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured: starts with purpose, then categories, then simple args/returns. Every sentence is informative. Could be slightly more concise (e.g., the args section is minimal), but overall efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (thousands of indicators with theme/search), the description provides adequate categories and filter options. The output is described as a Markdown list with IDs and descriptions, which aligns with the exploration purpose. No output schema was provided, but the description covers key aspects.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema descriptions cover all parameters (theme, search, limit) with clear details, so the description need not repeat them. However, the description adds value by listing example indicator categories (e.g., LR.*, NERA.*) that help the agent understand what to search for, going beyond the schema's formal descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly specifies 'listet verfügbare Indikatoren' (lists available indicators) from UNESCO Institute for Statistics, with a scope of over 4,000 indicators across education, science, culture. It distinguishes from sibling tools like uis_get_education_data (which fetches data) by focusing on exploration and search.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly states the function is for 'Exploration und Indikatorsuche' (exploration and indicator search), providing categories to guide use. It doesn't explicitly list when not to use or contrast with siblings, but the categories and purpose imply the correct context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Listet verfügbare Versionen der UNESCO UIS-Datenbank auf.
Die UIS veröffentlicht mehrmals jährlich neue Datenversionen. Nützlich um sicherzustellen, dass mit den neuesten Daten gearbeitet wird.
Returns: Markdown-Liste mit Versionsbezeichnungen und Publikationsdaten
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already provide readOnlyHint, destructiveHint, idempotentHint, and openWorldHint. The description adds value by explaining that UIS publishes multiple new versions per year and that the tool returns a Markdown list with version names and publication dates, which is beyond the annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is very concise with only three sentences, front-loaded with the core action. Every sentence adds value: stating function, providing context about publication frequency, and specifying the return format. No unnecessary text.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given zero parameters and a simple listing task, the description is complete. It explains the purpose, the context of version releases, and the return format (Markdown list). Annotations cover safety and idempotency, and an output schema exists. No gaps remain.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
No parameters exist, so schema coverage is 100% and baseline is 4. The description does not add parameter information because none are needed. This is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool lists available versions of the UNESCO UIS database, distinguishing it from sibling tools like uis_list_countries and uis_list_indicators. The verb 'listet' and specific resource 'Versionen der UNESCO UIS-Datenbank' provide clear purpose.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description mentions it is useful to ensure working with the latest data, implying usage context, but does not provide explicit when-to-use or when-not-to-use guidance or compare with alternatives. No exclusions are stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.3.0education_benchmark_countriesoecd_get_education_indicatoroecd_list_education_datasetsoecd_search_datasetsuis_compare_countriesuis_country_education_profileuis_get_education_datauis_list_countriesuis_list_indicatorsuis_list_versionsScored across 10 tools
Each tool targets a distinct operation: listing, searching, retrieving raw data, comparing countries, generating benchmarks, or creating profiles. The only potential overlap (education_benchmark_countries vs uis_compare_countries) is actually complementary (multi-indicator benchmark vs single-indicator comparison), so no ambiguity.
All tools follow a consistent <source>_<verb>_<noun> pattern in snake_case (e.g., uis_list_indicators, oecd_get_education_indicator). The prefix indicates the data source (education_, oecd_, uis_), making it easy to navigate.
With 10 tools, the server covers two major data sources (UNESCO UIS and OECD) across essential operations: discovery, retrieval, comparison, benchmarking, and country profiling. This is a well-scoped set that avoids bloat while providing sufficient functionality.
The tool surface covers the full lifecycle of accessing education statistics: listing available indicators/countries/versions, searching, fetching raw data, comparing countries, creating profiles, and running multi-country benchmarks. No obvious gaps for a read-only data API.
UNESCO UIS statistics (education, science, culture) with full provenance and fixed releases.
Give your agent web search and authoritative datasets: S&P Global, FRED, OECD, SimilarWeb & more.
Search and query 1,500+ OECD statistical datasets via SDMX. Keyless.
Macro data for AI agents: GDP, inflation, unemployment and more (World Bank, US BLS). No keys.
