feat: add agent marketplace swarm example with skill-based routing

[Yatiraj12]

Summary

Problem: The repository currently provides a multi-agent swarm example but does not demonstrate skill-based routing or an agent marketplace architecture where agents advertise capabilities and tasks are dynamically routed.

Why it matters: Developers building on Bindu need a clear example of agent discovery, skill registration, and modular orchestration to design scalable AI agent systems.

What changed: Added a new example examples/agent_marketplace_swarm demonstrating a dynamic agent marketplace with skill-based routing using a router agent, skill registry, and orchestrator coordinating specialized agents.

What did NOT change (scope boundary): No changes were made to Bindu core libraries, runtime behavior, APIs, storage backends, or scheduler components.

Change Type (select all that apply)

• Bug fix
• [] Feature
• Refactor
• [] Documentation
• Security hardening
• Tests
• Chore/infra

Scope (select all touched areas)

• Server / API endpoints
• Extensions (DID, x402, etc.)
• Storage backends
• Scheduler backends
• Observability / monitoring
• Authentication / authorization
• CLI / utilities
• Tests
• [] Documentation
• CI/CD / infra

Linked Issue/PR

• Closes #
• Related #

User-Visible / Behavior Changes

None.

Security Impact (required)

• New permissions/capabilities? (No)
• Secrets/credentials handling changed? (No)
• New/changed network calls? (No)
• Database schema/migration changes? (No)
• Authentication/authorization changes? (No)
• If any Yes, explain risk + mitigation:

Verification

Environment

• OS:Windows 11
• Python version:3.11
• Storage backend:Memory (default example config)
• Scheduler backend:Memory (default example config)

Steps to Test

1.Navigate to the example directory:

2.cd examples/agent_marketplace_swarm

3.Configure environment variables:
GROQ_API_KEY=your_api_key
MODEL_NAME=llama-3.3-70b-versatile

4.Run the agent:
python bindu_super_agent.py

5.Send a request such as:
Explain quantum computing

Expected Behavior

The request is routed through:

Router Agent → Skill Registry → Selected Agent → Response
Example:

User request: "Explain quantum computing"
Router selects → Research Agent
Research Agent generates response

Actual Behavior

The example successfully:
Registers agents dynamically
Routes tasks via the skill registry
Executes the correct specialized agent
Returns the LLM response.

Evidence (attach at least one)

• Failing test before + passing after
• [] Test output / logs
• [] Screenshot / recording
• Performance metrics (if relevant)

Human Verification (required)

What you personally verified (not just CI):

• Verified scenarios:
  Research query routed to Research Agent
  Translation requests routed to Translator Agent
  Summarization requests routed to Summarizer Agent
• Edge cases checked:
  Unknown request returns fallback response
  Empty input handled safely
• What you did NOT verify:
  Distributed agent communication
  Production deployment scenarios

Compatibility / Migration

• Backward compatible? (Yes)
• Config/env changes? (No)
• Database migration needed? (No)
• If yes, exact upgrade steps:

Failure Recovery (if this breaks)

• How to disable/revert this change quickly:
• Simply remove the examples/agent_marketplace_swarm directory.
• Files/config to restore:None (core system untouched)
• Known bad symptoms reviewers should watch for:
• None expected since this example does not modify runtime components.

Risks and Mitigations

None.

• Risk:
  • Mitigation:

Checklist

• [] Tests pass (uv run pytest)
• [] Pre-commit hooks pass (uv run pre-commit run --all-files)
• [] Documentation updated (if needed)
• [] Security impact assessed
• [] Human verification completed
• [] Backward compatibility considered

[Yatiraj12]

This example demonstrates a dynamic agent marketplace where agents register skills and are routed via a skill registry. It can serve as a reference implementation for building agent ecosystems with Bindu.

[Paraschamoli]

hey @Yatiraj12 use open router api key , don't make research agent it already exists in example also don't use that complex folder structure keep it simple

[Yatiraj12]

Thanks for the feedback!

I’ve updated the example to:

• use OpenRouter API instead of Groq
• remove the duplicate research agent
• simplify the structure and routing logic to align with existing examples

Please let me know if anything else needs improvement.

[Paraschamoli]

hey @Yatiraj12 agent is not working.

input:-Explain quantum computing

output:-<coroutine object Orchestrator.run at 0x000001943845B140> there might be async await problem or the agent are not coordinating so do changes and test the agent.