ERC3 Sample Agents

Sample agents demonstrating how to participate in the ERC3: AI Agents in Action competition.

Getting Started

1. Get Your API Key

To use these agents, you'll need an ERC3 API key:

1. Visit https://erc.timetoact-group.at/
2. Enter the email address you used during registration
3. Your API key will be displayed

Note: If you haven't registered yet, https://www.timetoact-group.at/events/enterprise-rag-challenge-part-3 and allow 24 hours for your registration to be processed.

2. Prerequisites

All agents require:

• ERC3 SDK - for connecting to the platform and accessing benchmarks
• ERC3_API_KEY - your competition API key
• LLM API Key - such as OPENAI_API_KEY or equivalent (depending on the agent)

Running an Agent

Here's an example of running the sgr-agent-store (a simple agent that solves the store benchmark):

# Set up your environment variables
export OPENAI_API_KEY=sk-...
export ERC3_API_KEY=key-...

# Navigate to the agent directory
cd sgr-agent-store

# Activate your virtual environment (optional but recommended)
# python3 -m venv venv
# source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the agent
python3 main.py

this should look like this:

you can also go to the Web UI and view your running session there:

Task execution view allows to see interactions of agent with the Benchmark in more detail:

Available Agents

• sgr-agent-store - A simple agent implementation for STORE benchmark. It relies on Schema-Guided Reasoning to provide adaptive thinking capabilities with a single recursive prompt and gpt-4o.
• sgr-agent-erc3 - A simple SGR NextStep agent for ERC3-DEV benchmark.

Resources

• https://www.timetoact-group.at/events/enterprise-rag-challenge-part-3
• https://erc.timetoact-group.at/

Support

You can ask questions in the discord channel (you get a link to that with the registration email)