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., "@timeweaver-mcp Generate 90 days of daily e-commerce sales using the ecommerce_sales preset."
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.
Synthetic time-series test data, on demand, inside your AI client. Generate realistic series with configurable trend, seasonality, noise, anomalies, and multiple correlated streams — perfect for testing dashboards, charts, monitoring/alerting, forecasting models, and anomaly detection. Output as JSON, CSV, or SQL.
Part of the fixturelab test-data tools. Its sibling SeedWeaver does relational/database test data.
LLMs are unreliable at hand-generating coherent time-series — trends drift, "seasonality" doesn't actually repeat, and correlations between series are fake. TimeWeaver generates data with verifiable statistical properties: a linear trend really has the slope you asked for, a seasonal cycle really repeats at its period, two correlated series really hit the target correlation, and AR(1) noise really has the autocorrelation you set.
Related MCP server: Faker MCP Server
npx -y timeweaver-mcpAdd to your MCP client config (e.g. Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"timeweaver": {
"command": "npx",
"args": ["-y", "timeweaver-mcp"]
}
}
}To unlock Pro, add your license key:
{
"mcpServers": {
"timeweaver": {
"command": "npx",
"args": ["-y", "timeweaver-mcp"],
"env": { "TIMEWEAVER_LICENSE": "YOUR-KEY-HERE" }
}
}
}generate_timeseries — generate data from a preset and/or explicit components (length, frequency, baseline, trend, seasonality, noise, anomalies, correlated series). Output JSON / CSV / SQL.
list_presets — list built-in presets: ecommerce_sales, server_cpu, iot_temperature, website_traffic, stock_price, api_latency_ms.
"Generate 90 days of daily e-commerce sales using the ecommerce_sales preset."
"Generate 3 correlated server CPU series over 500 minutes with correlation 0.8, as CSV."
"Make an hourly temperature series with a daily cycle and a level shift on day 5, as SQL into a table called readings."
Free | Pro | |
Points per series | 200 | up to 100,000 |
Series | 1 | up to many, correlated |
Trend | none / linear | + exponential, logistic |
Seasonality | 1 cycle | multiple cycles |
Noise | gaussian | + AR(1) autocorrelated |
Anomalies | – | spikes, level shifts, trend changes, dropouts |
Output | JSON | + CSV, SQL |
Deterministic seed | – | ✓ |
Pro: $19/mo or $39 one-time → https://fixtureforge.gumroad.com/l/timeweaver
MIT (the server code). Pro features require a valid license key.
Generate realistic synthetic time-series data with configurable trend, seasonality, noise, anomalies, and multiple correlated series. Ideal for testing dashboards, charts, monitoring/alerting, forecasting and anomaly-detection. Output as JSON, CSV, or SQL INSERTs. Use a preset for quick sensible defaults, or specify components explicitly.
Inputs: max, min, seed, names, +17 more.
List the built-in time-series presets (realistic ready-made configurations like e-commerce sales, server CPU, IoT temperature, website traffic, stock price, API latency). Use a preset name with generate_timeseries to get sensible defaults you can still override.
Inputs: no named parameters.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 2 tools
The two tools are completely distinct: one lists predefined presets, the other generates actual time-series data. There is no overlap or ambiguity between them.
Both tool names follow the consistent verb_noun pattern with snake_case: list_presets and generate_timeseries. The naming is predictable and clear.
With only 2 tools, the server feels minimal for the described functionality (configurable trend, seasonality, noise, anomalies, multiple output formats). While focused, it could benefit from splitting generation into separate tools for configuration or output format selection.
The server covers the core use case of generating time-series data with presets, but lacks tools for creating or editing presets, which would be a natural extension. The missing capability is minor and agents can work around it by overriding parameters.
Generate realistic, FK-consistent synthetic test data for your databases from your AI assistant.
Generate realistic relational test data — 156 field types, 22 locales, JSON/CSV/SQL, free previews.
Privacy-preserving synthetic health data generation. FHIR R4/R5 compliant.
Generate synthetic random user data for testing, demos, and development without using real persona.
