# EigenformAI/geocluster-mcp

Source: https://glimind.com/server/github%2Feigenformai%2Fgeocluster-mcp
Registry id: `github/eigenformai/geocluster-mcp`
Status: unknown

## Tools (37)

- `aggregate`: Aggregate by grid cell or by category.
- `align_grids`: Match a raster to a reference grid and CRS.
- `band_math`: Raster algebra with b1, b2, ... e.g. (b1-b2)/(b1+b2).
- `check_missing`: Columns with missing values.
- `clip_to_extent`: Clip a raster to a bounding box.
- `cluster`: k-means or DBSCAN clustering.
- `compute_anomaly`: Anomaly scores by z-score, MAD, or ratio.
- `compute_gradient`: Edge detection (Sobel).
- `convert_dtype`: Cast columns to numeric, int, float, str, or datetime.
- `detect_cleaning_issues`: Granular per-column diagnosis.
- `end_training_session`: End a training session.
- `evaluate_hypotheses`: Evaluate hypotheses.
- `export_artifact`: Bundle results as zip, xlsx, or geojson.
- `filter_rows`: Row filter (==, >, in, contains, not_null, ...).
- `fix_decimals`: Convert comma-decimal values to dot-decimal.
- `generate_hypotheses`: Generate hypotheses.
- `get_training_stats`: Get training statistics.
- `inspect_dataset`: Structural summary; loads dataset into cache for query_data.
- `inspect_raster`: Raster shape, CRS, bounds, transform, per-band stats.
- `inspect_specific_columns`: Per-column stats, unique values, or value counts.
- `list_files`: List files in a workspace directory.
- `merge_datasets`: Join two datasets.
- `parse_detection_limits`: Turn <0.01 style strings into numbers.
- `plot_map`: Map from a raster or an x/y CSV.
- `profile_geochem`: Depth-weighted mean, max, and grade-thickness for assay intervals.
- `query_data`: Filter, sort, threshold, or evaluate expression against cached dataset.
- `rank_by_metric`: Top N rows by a metric.
- `reduce_dimensions`: PCA or UMAP dimensionality reduction.
- `reproject`: Reproject a raster or CSV of x/y points to a new CRS.
- `resample`: Change raster resolution (0.5 halves, 2.0 doubles).
- `select_bands`: Extract raster bands (1-based).
- `start_training_session`: Start a training session.
- `summarize_provenance`: List every artifact in results/ with metadata.
- `texture_features`: Texture rasters (entropy).
- `threshold`: Keep rows above or below a cutoff.
- `validate_geology`: Scan dataset for common quality issues.
- `verify_claims`: Check numeric claims against the source data.

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.