For Agents
Drive guided interviews and document assembly on a docassemble server, manage users and files, and automate intake-to-document pipelines.
Get started with Docassemble API in minutes using your preferred integration method.
# Add to your MCP client config (Claude Desktop, Cursor, Windsurf)
{
"jentic": {
"url": "https://api.jentic.com/mcp",
"auth": "oauth"
}
}
# Then ask your agent:
"start a docassemble interview session"
# → Jentic returns the GET /events tool with parameter schema, agent executes.What an agent can do with Docassemble API API.
Start a new interview session from a YAML interview file
Step through interview questions, posting answers to advance the session
Go back to a previous interview screen or trigger a defined action
Upload and download files attached to an interview session
List and create users on the docassemble server
GET STARTED
Use for: Start a new docassemble interview for a tenant rights intake, Post answers to advance an in-progress interview to the next question, Go back to the previous question in an active session, Trigger a custom action defined in the interview
Not supported: Does not handle e-signature, court e-filing, payment processing, or general-purpose word-processor editing — use for guided interview and document assembly only.
Docassemble is an open-source platform for guided interviews that assemble legal documents and intake forms. Its API exposes 25 endpoints for starting and stepping through interview sessions, managing users, files, packages, the playground and server configuration. Authenticated via the X-API-Key header, it lets external apps embed legal-tech intake flows, automate document generation pipelines, and synchronise users with a host docassemble server.
Pull and update server configuration and packages
Use the playground endpoints for development workflows
Patterns agents use Docassemble API API for, with concrete tasks.
★ Embedded Legal Intake
A legal aid organisation can embed a docassemble interview into its website and step the user through it via the API rather than redirecting to the docassemble UI. POST /api/session/new starts the interview, GET /api/session/question reads the next question, and POST /api/session submits answers. The result is a branded intake flow that still benefits from the underlying interview logic.
POST /api/session/new with the interview filename, then loop calling /api/session/question and POST /api/session until completion.
Document Assembly Pipeline
Use the API to script document generation in batch — for example, generating tenant notices, immigration forms or contracts from CSV input. A worker process starts a session per row, posts the answers, and retrieves the assembled PDF from /api/file/{file_number}. Cuts manual drafting time on repetitive document workflows.
For each row in a CSV, start a new session, post answers, then call /api/file/{file_number} to download the resulting PDF and save it under the row id.
User and Server Administration
Operations teams running a multi-tenant docassemble server can automate user provisioning, package updates and configuration changes via the API. Endpoints under /api/user, /api/user_list and the configuration routes let an admin script bulk-add users from an HR feed or roll a new package version across instances.
Iterate the HR export and POST /api/user/new for each new hire, then GET /api/user_list to verify the count matches.
AI Agent Legal Tech Assistant
Through Jentic, an AI assistant can guide a user through a docassemble interview by translating natural-language replies into the structured answers the API expects. The agent searches Jentic for the interview intent, loads the session operations, and posts answers back. This combines docassemble's structured legal logic with conversational UX.
Use Jentic to search 'docassemble session question', load /api/session/question, and execute it after each user reply, posting answers back to /api/session.
25 endpoints — docassemble is an open-source platform for guided interviews that assemble legal documents and intake forms.
METHOD
PATH
DESCRIPTION
/api/session/new
Start a new interview session
/api/session/question
Get the current question for a session
/api/session
Submit answers to advance the session
/api/session/back
Go back to the previous question
/api/session/action
Trigger a defined interview action
/api/interviews
List interviews on the server
/api/file/{file_number}
Download an assembled file
/api/user/new
Create a new user on the server
/api/session/new
Start a new interview session
/api/session/question
Get the current question for a session
/api/session
Submit answers to advance the session
/api/session/back
Go back to the previous question
/api/session/action
Trigger a defined interview action
Three things that make agents converge on Jentic-routed access.
Credential isolation
Docassemble X-API-Key values are stored encrypted in the Jentic vault. Agents receive scoped access via Jentic's MAXsystem and never see the raw key, which matters for self-hosted servers handling sensitive intake data.
Intent-based discovery
Agents search by intent (for example 'start a docassemble interview' or 'submit interview answers') and Jentic returns the matching session operations such as /api/session/new and /api/session with their input schemas, so the agent steps through interviews without parsing docs.
Time to first call
Direct Docassemble integration: 2-3 days to wire session lifecycle, file download and error handling. Through Jentic: under 1 hour — search, load schema, execute.
Alternatives and complements available in the Jentic catalogue.
Open edX LMS API
Open-source education platform — pairs with docassemble for legal-education interview content
Use Open edX when the task is course delivery rather than guided document interviews.
Concord
Commercial contract lifecycle management platform — alternative when teams need e-signature and post-execution tracking on assembled documents
Choose Concord for a full CLM with e-signature; choose docassemble for guided interviews that build the document up front.
GitHub REST API
Manage docassemble interviews stored as GitHub repositories.
Use GitHub to version-control docassemble package code.
Twilio API
Trigger SMS reminders from a docassemble interview workflow.
Pair Twilio when SMS or voice prompts need to accompany a legal interview.
Specific to using Docassemble API API through Jentic.
What authentication does the Docassemble API use?
Docassemble uses an API key in the X-API-Key header. Keys are issued per user on the host docassemble server. When called through Jentic, the key lives in the Jentic vault and the agent receives only a scoped reference.
Can I drive an entire interview through the Docassemble API?
Yes. Start the session with POST /api/session/new, then loop calling GET /api/session/question to read the next prompt and POST /api/session with the answers. /api/session/back and /api/session/action handle navigation and custom triggers.
What are the rate limits for the Docassemble API?
Docassemble is typically self-hosted, so rate limits are governed by the host server's resources rather than a vendor quota. Treat heavy batch jobs carefully and watch CPU on the docassemble worker nodes.
How do I retrieve an assembled document with the Docassemble API through Jentic?
After completing an interview, call GET /api/file/{file_number} to download the assembled PDF or DOCX. Through Jentic, search 'docassemble download file', load the operation and execute it with the file number returned by the session.
Can I create users on a Docassemble server with the API?
Yes. POST /api/user/new creates a new user account on the server, and GET /api/user_list returns the current roster, which makes it straightforward to sync users from an HR or directory system.
Is the Docassemble API free?
Docassemble itself is open source and free to self-host, so API access has no vendor cost. Your only costs are the server hosting and any commercial extensions you install on top.
/api/interviews
List interviews on the server
/api/file/{file_number}
Download an assembled file
/api/user/new
Create a new user on the server