Discover & Connect
Once Robot understands a request, it can determine whether other people, services, capabilities, or agents may be needed.
This is where discovery becomes important.
From a Request to a Searchable Requirement
Suppose the user asks:
“Find a BUET teacher for a Class 9 student who can teach four days a week for ৳5,000 per month.”
Robot can turn that natural-language request into a set of requirements:
- Purpose: Teaching
- Student: Class 9
- Qualification: BUET
- Frequency: Four days per week
- Budget: ৳5,000 per month
These requirements can then be used to identify potentially relevant participants.
Agent Discovery
A participant in the Craftly ecosystem may have an associated agent that can represent their available capabilities or services.
The discovery process can therefore look like:
↓
Requirement Analysis
↓
Find Relevant Agents
↓
Check Capabilities
↓
Contact Suitable Agents
The goal is not simply to produce a list of names.
The goal is to find agents that are relevant to the actual objective.
In broader agent interoperability work, capability discovery is a foundational concept: agents can publish metadata describing their identity, skills, endpoints, and interaction requirements so other agents can determine whether they are suitable for a task.
Craftly Robot applies the same general idea to its own ecosystem and identity model.
Connect
After identifying potentially relevant participants, Robot can initiate agent-to-agent communication.
This can allow agents to exchange the information necessary to determine whether a potential match exists.
Depending on the task, that information may involve:
- Availability
- Capabilities
- Requirements
- Scheduling
- Pricing
- Service conditions
- Other task-specific constraints
Only information relevant to the task should be exposed, subject to the identity, privacy, and permission rules of the participating agents.