Skip to content
Back to search

geosql

MarketplaceWorks with Claude CodeGitHub

GeoSQL plugin by Dekart: geospatial SQL and interactive maps for Claude.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/dekart-xyz--geosql/badge.svg)](/listings/dekart-xyz--geosql)

HTML

<a href="/listings/dekart-xyz--geosql"><img src="/api/listings/dekart-xyz--geosql/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

With Python (interactive mode):

pip install geosql && geosql

GeoSQL first installs the skill into your detected agent. It then checks whether the optional Dekart CLI is ready and, only when needed, asks before installing or initializing it. The prompt shows the exact commands and defaults to the recommended install option.

After installing, prompt your agent like this:

/geosql Show EV charger density along major roads and render a map

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 19 lines
  1. {
  2. "name": "geosql",
  3. "description": "GeoSQL plugin by Dekart: geospatial SQL and interactive maps for Claude.",
  4. "owner": {
  5. "name": "Dekart XYZ",
  6. "email": "support@dekart.xyz",
  7. "url": "https://dekart.xyz"
  8. },
  9. "plugins": [
  10. {
  11. "name": "geosql",
  12. "source": "./",
  13. "description": "Geospatial SQL and maps for BigQuery, Snowflake, DuckDB, Wherobots and Postgres: discovers Overture Maps and warehouse schemas, writes cost-safe SQL with a dry-run budget check, validates results and renders them on an interactive Dekart map through the GeoSQL MCP tools or the dekart CLI.",
  14. "category": "location",
  15. "homepage": "https://github.com/dekart-xyz/geosql",
  16. "version": "0.39.0"
  17. }
  18. ]
  19. }
README.md · README · 106 lines
  1. # GeoSQL
  2. Claude Code, Codex, GitHub Copilot, OpenCode, and Mistral Vibe skill for data scientists and analysts working with geospatial data on PostGIS, BigQuery, Snowflake, and Wherobots.
  3. > Note: No SaaS account needed. Works 100% locally or self-hosted.
  4. ![GeoSQL demo](assets/geosqldemo.gif)
  5. > 4x improvement on geospatial tasks with map in the loop.
  6. ![Agent with maps loop, 4x performance](assets/agent-with-maps-4x.png)
  7. ## Quick Start
  8. With Python (interactive mode):
  9. ```bash
  10. pip install geosql && geosql
  11. ```
  12. GeoSQL first installs the skill into your detected agent. It then checks whether the optional Dekart CLI is ready and, only when needed, asks before installing or initializing it. The prompt shows the exact commands and defaults to the recommended install option.
  13. After installing, prompt your agent like this:
  14. ```
  15. /geosql Show EV charger density along major roads and render a map
  16. ```
  17. ## How to update
  18. Upgrade GeoSQL in the Python environment where it is installed:
  19. ```bash
  20. pip install --upgrade geosql
  21. ```
  22. If you use the optional Dekart CLI, update it too:
  23. ```bash
  24. pip install --upgrade dekart
  25. ```
  26. Then rerun the installer so the latest GeoSQL skill and references are copied
  27. to your detected agent:
  28. ```bash
  29. geosql
  30. ```
  31. ## Version telemetry
  32. GeoSQL sends a best-effort version check with its package version and an opaque random installation ID. GeoSQL and Dekart CLI share the ID at `${XDG_CONFIG_HOME:-~/.config}/dekart/installation_id`; deleting that file resets it. Set `DO_NOT_TRACK=1` (or `DNT=1`) to disable the version check before an ID is created. CI sends a reserved non-persisted test ID.
  33. ## Example prompts to try in your agent:
  34. Real estate analysis:
  35. ```
  36. /geosql Show buildings with low school accessibility in Ottawa, render as a map
  37. ```
  38. Site selection:
  39. ```
  40. /geosql Find the top 10 locations for Sporting Goods Store in Seattle based on POI co-location and distance to the nearest competitor. Create a map.
  41. ```
  42. EV charging infrastructure:
  43. ```
  44. /geosql create map EV charger density along major Romanian roads, highlighting how many charging stations are within 5 km of each motorway, trunk, or primary road segment.
  45. ```
  46. ## How it works
  47. GeoSQL runs an agent loop with a map in it.
  48. ![How GeoSQL connects the agent, Dekart CLI, data sources, and deployment](assets/how-geosql-works.png)
  49. 1. **Discovery.** The skill explores your warehouse metadata (tables, columns, types) instead of guessing schemas. Works with Overture Maps shares on BigQuery and Snowflake, and your private tables on PostGIS, BigQuery, Snowflake, or Wherobots.
  50. 2. **SQL.** The agent writes spatial SQL using the right functions for your engine (`ST_INTERSECTS`, `ST_DISTANCE`, H3, bbox overlap for partition pruning, and so on).
  51. 3. **Cost check.** On BigQuery, every query is dry-run first to estimate bytes scanned. A 10 GiB billing cap is enforced by default. Over-budget queries get rewritten cheaper (tighter bbox, lower H3 resolution, more filters) instead of executed.
  52. 4. **Geometry validation.** The agent computes total area (polygons) or total length (lines) as a sanity check, and cross-checks against domain knowledge.
  53. 5. **Map feedback.** When available, the agent renders the result through Dekart, looks at the rendered image, and corrects geometry mistakes the text-only loop would miss. This is the loop that gets the 4x improvement.
  54. The skill uses your local CLI authentication (`bq`, `snow`, `dekart`), so warehouse credentials never go to the agent.
  55. ## Benchmarks
  56. GeoSQL ships with a reproducible eval suite under `evals/`. Each case asserts specific behaviors (cost guardrails, validation steps, correct result), not just "did the agent answer."
  57. Current results on the included suite:
  58. | Case | Assertions | Pass rate |
  59. | --- | --- | --- |
  60. | `london-boroughs` | 4 | 100% |
  61. | `berlin-create-map` | 3 | 100% |
  62. | `paris-boundaries` | 1 | 100% |
  63. | **Total** | **8** | **100%** |
  64. Average: 3,085 tokens per turn, 72 s duration per turn.
  65. The **4x improvement** chart above compares the same task set with and without the map-in-loop step. Without maps, the agent's text-only validation misses geometry-class errors (mistaking a neighborhood polygon for a metro-area perimeter, double-counting overlapping features, picking the wrong join key on coordinate-reference systems). Adding the rendered map as a tool call lets the agent see those mistakes and self-correct.
  66. Run the suite yourself:
  67. ```bash
  68. python evals/run.py
  69. ```
  70. See `evals/RUNBOOK.md` for setup and how to add new cases. PRs with new evals welcome.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.