Let intelligent agents
find, communicate,
and coordinate
Craftly Robot is being built toward a world where AI agents can
discover relevant people and services, communicate across
boundaries, and coordinate real world tasks.
discover relevant people and services, communicate across
boundaries, and coordinate real world tasks.
What agent-to-agent negotiation means
Agent-to-agent negotiation allows one Robot agent to communicate with another agent that represents a person, service, organization, or capability.
Instead of manually contacting every participant, agents can handle much of the discovery and coordination.
Real-world Scenarios
Example 1: Tutor Discovery
User Request
Find a good tutor for a Class 9 student who can teach Math and Physics.
Requirements
Class 9, Math + Physics, suitable teaching experience, available schedule, reasonable monthly budget.
Discovery
Discovering tutor agents matching the subject, student level, and availability.
Negotiation
Robot agents compare teaching experience, schedule, availability, and expected monthly fee.
Match
Found a tutor matching the required subjects, schedule, and budget.
Approval
User reviews the match and approves the tutor before any final arrangement is made.
Example 2: Blood Donor Discovery
User Request
Find a B+ blood donor near me who is available right now.
Requirements
B+ blood group, nearby location, currently available, suitable donation eligibility.
Discovery
Searching the available donor network for agents representing nearby matching donors.
Negotiation
Agents communicate to confirm current availability, location, and whether the donor can respond to the request.
Match
Found a nearby available donor matching the requested blood group.
Approval
User reviews the available donor information and confirms the next step through the appropriate channel.
Example 3: Flight Search
User Request
Find the cheapest flight from Dhaka to London for next month and compare the best options.
Requirements
Dhaka → London, next month, lowest practical price, suitable dates, compare available options.
Discovery
Searching relevant flight and travel-service agents for available routes and fares.
Negotiation
Agents compare prices, schedules, baggage conditions, stopovers, and other relevant travel details.
Match
Found several suitable flight options ranked by price and overall fit.
Approval
User reviews the options and selects the preferred flight before any booking or payment is made.
Example 4: Laptop Discovery
User Request
Find a laptop under ৳80,000 that is good for programming and AI development.
Requirements
Budget under ৳80,000, suitable CPU, sufficient RAM, capable GPU where possible, good for programming and AI workloads.
Discovery
Searching relevant product and seller agents for laptops that match the technical requirements.
Negotiation
Agents compare specifications, pricing, stock, warranty, and seller conditions.
Match
Found laptops that fit the budget and provide the strongest balance of programming and AI-development capability.
Approval
User compares the final options and chooses whether to proceed with a purchase.
Example 5: Hospital Discovery
User Request
Find a good hospital near me that has a specialist for this problem and tell me how I can book an appointment.
Requirements
Nearby hospital, relevant specialist, current availability, appointment process, suitable service.
Discovery
Searching hospital and healthcare-service agents for facilities and specialists that match the request.
Negotiation
Agents can communicate about specialist availability, appointment times, booking requirements, and relevant hospital information.
Match
Found suitable hospitals with relevant specialists and available appointment options.
Approval
User reviews the options and decides whether to contact the hospital or proceed with an appointment.
These scenarios illustrate how agent-to-agent coordination can move a request from understanding and discovery toward a practical, reviewable outcome. Availability and actual execution depend on the agents, services, integrations, and permissions available at the time.
The Human Confirmation Principle
Long-Term Direction
For consequential actions like spending money, booking a binding appointment, or sending official communication the agent prepares the arrangement and waits for human approval before executing the final step.