How to activate multiple agents at once in Claude Code for a coordinated workflow?

[BittuChan]

I've got the agency-agents installed in ~/.claude/agents/ and individual agents work great, but I'm trying to figure out how to activate multiple agents for a single task.

For example, I want the Frontend Developer, Security Engineer, and QA Tester to all collaborate on reviewing a React component. Right now I'm doing:
"Activate Frontend Developer and review this component"

Then separately:
"Now activate Security Engineer and check for vulnerabilities"

Is there a way to activate multiple agents simultaneously so they coordinate their feedback instead of me manually chaining them?

[Shiven0504]

hey @BittuChan, Claude Code doesn't support multi-agent orchestration natively, each session runs one agent at a time. Your sequential approach is currently the right pattern.
I can suggest a workaround for coordinated reviews:

Create a meta-prompt that references multiple agent perspectives:

Review this component from three perspectives:
1. Frontend Developer (code quality, React best practices)
2. Security Engineer (XSS, injection risks, auth)
3. QA Tester (edge cases, error states)

Provide consolidated feedback.

This simulates multi-agent collaboration within a single Claude session.

[Shiven0504]

Alternatively, the examples/ directory in the repo shows multi-agent outputs, but those were generated by running agents separately and combining results manually - not through automated orchestration.

For true parallel execution, you'd need a custom wrapper script that invokes Claude Code multiple times :)

[BittuChan]

@Shiven0504, Yeah it makes sense. The meta-prompt approach is working well enough for my use case. Thanks for clarifying that orchestration isn't built-in!