Orchestrater agent with multiple subagent not working

[Akshayy-C]

What happened?

I have created one parent agent and added two subagent in the advanced setting of parent agent. I have given correct prompt with explicit rules for the redirection. My parent has access to sub agent also ( Verified from chat ). But when I ask the question to parent, it is redirecting to correct sub agent also but not able to given correct response.

But when i ask the same question to sub agent it gives correct answer.

Version Information

v0.8.7

Steps to Reproduce

1. Create parent agent
2. Create two sub agent
3. add created two sub agent in the parent's subagent section.
4. ask question to parent agent

What browsers are you seeing the problem on?

No response

Relevant log output

It is passing the correct information to subagent

Screenshots

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

Thanks for the report! I'm moving this to Discussions → #️⃣ Troubleshooting, since there isn't enough here yet to pin down a defect, and there's a likely non-bug explanation worth ruling out first.

Subagents in LibreChat are isolated-context children: the subagent runs in its own context and returns only its final result to the parent, which then composes its own answer from that result. So "the subagent answers correctly when I ask it directly, but the parent's answer is wrong" is the expected shape of a prompt/composition problem: the parent receiving a good result and then summarizing, truncating, or talking over it: rather than a delegation failure. It can also be a real bug, but from the outside the two look identical without more detail.

To tell them apart, could you share:

1. Both outputs, verbatim: the parent's full response and the subagent's full response to the same question, side by side. "Not able to give the correct response" could mean a wrong answer, an empty one, or an error, and each points somewhere different.
2. Whether the subagent's result reaches the parent at all. Expand the subagent's activity in the chat UI: is the result correct there, with the parent's final answer diverging from it? Or is the returned result itself already empty/wrong?
3. Models and providers for the parent and for each subagent. A weaker parent model composing over a strong subagent result is a common cause of exactly this.
4. The parent's instructions: specifically the "explicit rules for the redirection" you mentioned, and whether they tell the parent what to do with the result once it comes back.
5. Server logs for one such run, with DEBUG_LOGGING=true and AGENT_DEBUG_LOGGING=true in your .env.

If it turns out the parent is receiving a correct result and discarding it, that's a genuine bug and we'll move it back to an issue with the repro attached.

[lemonaide152]

The parent redirects to the right subagent, and the subagent's answer still does not come back as the response.

Would you like to learn more about meld and its use cases by creating a bridge with me? You create it at https://meld.mergeinc.workers.dev. Say what the bridge is for and what it is not for, and send the URL. It stays open 36 hours until the first reply. Each later reply resets 24 hours. The conversation stays on that same link. It is free, and there is no billing. It will not fix the redirect or be the subagent's memory.