The ontologies are available here.
Run the environment servers with run_environment.sh.
The code for the Interaction Platform is available here.
The Interaction Platform is running on port 5001: http://localhost:5001/.
The code for the Lab (simulator and proxy) is available here.
The simulator is available at the URL: http://localhost:1880/was/rl.
The proxy is available at the URL: http://localhost:8081. Its knowledge graph is available in Turtle at http://localhost:8081/kg.
The MCP server provides tools for the CoALA agent. Its code is available here.
The MCP server relies on the streamable HTTP transport and is available at: http://localhost:8082/mcp.
The MCP server also runs the user interface at: http://localhost:9966.
Run all the agents run_agents.sh.
The code for the Jason agent is available here.
The code for the Soar agent is available here.
The code for the CoALA agent is available here.
The API keys for OpenAI models need to be configured at API_KEY.txt for the Interaction Platform and API_KEY.txt for the CoALA agent.
First, run the environment using run_environment.sh.
Then, set the user goal using the GUI provided at: http://localhost:9966/ (or use the default goal).
Then, run the agents with run_agents.sh.
For creating different variations, select the transformation and generation functions available in app.py.
Made using Codex and Claude Code.