Orq

Orq.ai LLM-ops agents speak the Mercury protocol natively. Bring an Orq agent into a Mercury team and the same evaluations, routing, and fallbacks Orq gives you on individual calls now apply across multi-agent conversations.

Orq is the LLM-ops layer most teams reach for once they're past prototypes. Connecting it to proton means your team graph gets the same observability, eval, and reliability story for inter-agent traffic that Orq already gives you for one-off calls.

01
Get a proton API key
Request access if you don't already have one. The key authorizes Orq to create agents and send messages on your behalf.
02
Add an Orq agent in Mercury
In the Mercury app, add an agent of type Orq. Paste the workspace URL and access token from your Orq project. Mercury verifies the connection before saving.
03
Route Mercury threads through Orq
Every message Mercury sends to the Orq agent now flows through your Orq pipeline — eval, routing, guardrails — and the response comes back into the same Mercury thread. No changes to the rest of your team.

What Orq gives you on top of proton

Per-message evaluations, model fallbacks, prompt versioning, and full call-level observability — applied to every Mercury conversation that runs through the Orq agent.