More screenshots
Works with the AI Agents you already use
geosense agent
Technical blueprint for a multi-phase geospatial AI agent with PostGIS, ML site-scoring, and 3D digital twins.
$2400
geosense agent
Example session with this skill installed
How do I set up the GeoSense Agent to use Google Gemini instead of Anthropic, and what are the specific version requirements for the database?
- Read your context and instructions
- Compiled the geosense agent
To switch the AI provider and verify your database compatibility, follow these configuration steps and version checks.
Provider Switch and Database Requirements
To change the AI provider to Google Gemini, modify your .env file as follows:
# .env configuration
AGENT_PROVIDER=gemini
GOOGLE_API_KEY=your_gemini_api_key_here
# Optional: specify the model
# GEMINI_MODEL_NAME=gemini-1.5-pro
Important: After updating the .env file, you must restart all services. Use stop_all.bat followed by start_all.bat (or the equivalent .sh scripts on Linux) to ensure the new environment variables are loaded into the LangGraph processes.
Required Database Versions
The project supports two primary configurations for PostgreSQL/PostGIS:
| Component | Project 1 Requirement | Project 2 Requirement |
|---|---|---|
| PostgreSQL | Version 18 | Version 16 |
| PostGIS | Version 3.6 | Version 3.4 |
| Extensions | postgis, postgis_raster | postgis, postgis_raster |
Next steps
- Run
pip install langchain-google-genaito ensure the Gemini connector is available. - Execute
deploy/create_readonly_role.sqlon your database to set up the required low-privilege API role. - Check the
/healthendpoint on port 8000 to verify the API recognizes the new provider.
Connects securely to your tools. The creator never sees your data.
About this skill
Building geospatial AI agents often leads to a mess of misaligned PostGIS schemas, failing LiDAR pipelines, and LLM hallucinations regarding location data. This skill provides the complete technical blueprint for the GeoSense Agent system, a production-ready LangGraph framework for geospatial Q&A and site-suitability analysis.
What it does
- Geospatial Agent Orchestration Coordinates 7 specialized tools via LangGraph and Claude/Gemini to answer complex location-based queries.
- Site-Suitability ML Executes a Phase 1 pipeline using OSM and DEM data in PostGIS to train RandomForest/XGBoost models for land scoring.
- Multi-Source Data Integration Manages workflows for Sentinel-2 land-cover classification, LiDAR building height analysis, and Mapillary street-view walkability.
- MCP & Digital Twin Support Configures a FastAPI/MCP server for tool access and a CesiumJS/Vite 3D digital twin for visualization.
- Production-Ready Deployment Defines Docker Compose configurations, GitHub Actions with keyless login, and hardened security controls for Cloud Run.
How it works
- Infrastructure Setup Initialize the project structure using provided PowerShell or Bash scripts and configure the Node 22 and Python 3.11 environments.
- Data Pipeline Register a city in
cities.json, load DEM and OSM extracts, and run therun_pipelinemodule to generate features and train the ML model. - Service Orchestration Launch the FastAPI backend, Streamlit interface, and 3D twin using the
start_allutility. - Agent Integration Select the AI provider (Anthropic or Gemini) in the environment configuration to enable the LangGraph reasoning loop.
Frameworks & tools
The system requires Python 3.11, Node 22 LTS, and PostGIS (v16 or v18). It utilizes LangGraph for orchestration, Scikit-learn/XGBoost for ML, and FastAPI for the MCP server. Frontend components use Vite, CesiumJS, and Streamlit.
Why this beats prompting it yourself
Prompting an LLM to handle GIS data results in coordinate errors and invented statistics. This skill enforces a "ground truth" workflow where the agent only reports results returned by verified PostGIS queries and ML model outputs. It solves the complex dependency chain between spatial databases, raster processing, and agentic tool-calling.
Use cases
- Site Selection Identify candidate development plots based on road proximity, hospital access, and flood risk.
- Risk Assessment Generate climate and terrain-based flood risk profiles for specific city coordinates.
- Urban Analysis Conduct digital twin reviews of building heights and canopy cover using LiDAR point clouds.
- Infrastructure Audits Evaluate street-level walkability using automated street-view image analysis.
Known limitations
This is a documentation and workflow skill, not an executable binary. The Linux conda environment file is untested. Site scoring is restricted to one trained city at a time.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Requires manual setup of PostGIS and specific versions of Node (22) and Python (3.11). Linux scripts are provided but untested. Site scoring model supports only one city per training run.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 day ago
- Software
Node.js