Discovery Response Agents
Last updated: October 23, 2026
Overview
A discovery response agent is like a drafting agent built for one job: answering written discovery. You give it the requests, and it returns a structured, formatted response in your firm’s voice, ready for your review.
Some common types of written discovery this agent can answer:
Interrogatories
Requests for production
Requests for admission
Discovery response agents live in the Agents section of Eve and run inside a matter, against the documents in that matter. Every workspace comes with example agents already built, and firms add their own.
How Discovery Response Agents Work
Discovery responses are repetitive in structure and different in substance. The format barely changes from case to case. The facts change every time. A discovery response agent keeps the formatting consistent so your attention goes to the second.
It learns your firm’s response format, language, and structure from the example documents you provide when you create it. After that, it lives in your firm’s library and works on any matter. You can generate as many responses from one agent as you need.
Two things carry over to every run:
The format. Your numbering, headings, and the way you state each response follow your examples.
The language. Your phrasing and tone follow your examples too, so responses read like your firm wrote them.
What changes with each run is the matter context, which is where the actual answers come from.
How They Differ From Other Agents
Chat agent | Drafting agent | Discovery response agent | |
|---|---|---|---|
Returns | A conversational response | A formatted document | A structured response to written discovery |
Best for | Analysis, summaries, recommendations | Demand letters, complaints, mediation statements | Interrogatories, requests for production, requests for admission |
Learns from | Saved instructions | Example documents | Example discovery responses |
A discovery response agent is a drafting agent with a narrower focus. If you’ve used drafting agents, this will feel familiar.
Where Discovery Response Agents Live
Open Agents from the left navigation, then See all agents for the full list. Discovery response agents sit alongside chat, drafting, auto, and communication agents. Use the Agent Type filter to see just the discovery response agents.
Each agent is either private to the person who created it or shared with the firm. Sharing is how one person’s well-built response format becomes the way the whole firm answers discovery.
When a Discovery Response Agent Is the Right Tool
If you’ve answered a set of interrogatories more than once, and the structure was basically the same each time, that’s a discovery response agent.
Some situations call for something else:
A one-off question about a single request. Ask in chat.
A document that isn’t a response to discovery. That’s a drafting agent.
A response format that’s different every time. If you’d have to rebuild the agent for each matter, there’s nothing to save.
If your firm answers discovery differently by case type or jurisdiction, create a separate agent for each format so each stays consistent.
What Context Does
Like every agent, a discovery response agent reads only the matter context selected for the run. A response can only draw on documents that were uploaded to the matter and left checked. If the answer to an interrogatory lives in a record that was never uploaded, the agent can’t find it.
You can also add firm reference material from the Central Library, such as standing instructions, alongside the matter context.
What They Can’t Do
They don’t file or serve anything. The output is a first draft for your review.
They don’t replace your judgment on what to answer. Check each response against the citations to the source documents before anything leaves the firm.
They only know what’s in context. Unchecked or missing documents can’t be read or cited.
They don’t contact anyone. Anything that goes to a client or outside party is a communication agent, which carries its own approval step.
Manual or Automatic
Like any agent, a discovery response agent can be run manually, set up as an auto agent, or both. Setting a trigger doesn’t remove the option to run it by hand. See the Auto Agents and Triggers article.
Key Terms
Discovery response agent: A drafting agent tuned for structured responses to written discovery.
Interrogatories: Written questions that must be answered in writing.
Requests for production: Requests for documents or other materials.
Requests for admission: Statements the other side asks you to admit or deny.
Example documents: The completed responses you provide when creating the agent, which teach Eve your firm’s format and language.
Matter context: The documents Eve reads during a run. Unchecked files are excluded entirely.
Central Library: Firm-wide reference material an agent can read alongside matter context.