Document Types
Every document carries its own intelligence. (Coming Soon)
The .aid format embeds prompt containers, context parameters, and file type designations directly into each document, so the AI knows what to do before you ask.
Documents should know what they are.
Each .aid file includes prompt containers that tell agents how to process the document, context parameters that adapt AI behavior to the matter, and file type designations (master files, sub-files, and training datasets) that classify the document within your firm's hierarchy. When an agent opens the document, it already knows the guidelines.
Prompt Containers
Embedded instructions that drive AI behavior
Prompt containers live inside the document itself. They specify how each section should be drafted, what context the inference engine should consider, and which compliance checks to run automatically. When an agent processes the document, it reads these containers first; no separate configuration needed. Containers can be scoped to specific sections, clauses, or the entire document.
The document tells the AI what to do. Not the other way around.
Master Files
Primary agreements that anchor a matter and define the document hierarchy
Sub-Files
Exhibits, schedules, and supporting documents that attach to and inherit context from master files
Training Datasets
Curated document collections that feed dynamic training and RAG preprocessing pipelines
Context Parameters
Variables embedded in documents that adapt AI output to jurisdiction, matter type, and client preferences
Classify
Master file, sub-file, or training dataset
Embed
Add prompt containers and context parameters
Execute
Agents follow embedded instructions automatically
.aid Format
A document format built for AI-native workflows
The .aid format extends standard document formats with a metadata layer purpose-built for AI interaction. It stores provenance tracking: which model generated each section, what context was used, what confidence level the engine assigned. It preserves the full reasoning chain so reviewers can audit not just what the AI wrote, but why. Documents carry their processing history, enabling cross-document learning and continuous improvement of your firm's AI capabilities.
The format that remembers everything the AI did, and why.
Provenance tracking
Every AI-generated section carries metadata: which model produced it, what context was assembled, what confidence score was assigned. Reviewers see exactly how each piece of content was created, enabling full auditability.
Document hierarchy
Master files, sub-files, and training datasets form a structured hierarchy. Sub-files inherit context from their parent master file, so exhibits and schedules automatically reflect the terms and definitions of the primary agreement.
Cross-document learning
The .aid format enables agents to learn across documents within a matter. Patterns from one agreement inform drafting of related documents, so the second exhibit benefits from what the AI learned processing the first.
Portable format
The .aid format exports cleanly to .docx and standard formats. Embedded intelligence stays with the document inside JR3, but clients and opposing counsel receive clean, standard files with no proprietary artifacts.
Coming Soon
Document type marketplace
Access pre-built .aid document types contributed by recognized specialists, including employment attorneys, M&A counsel, IP litigators, and compliance experts. Each document type comes with prompt containers, context parameters, and file hierarchies already configured. Use their frameworks directly inside Junior, or contribute your own to reach thousands of legal teams.
Pre-configured prompt containers
Specialist document types
Updated regularly
Interested in contributing?
We're actively seeking recognized specialists to contribute .aid document types for the marketplace. Package your expertise as intelligent document frameworks, complete with prompt containers and context parameters, and earn revenue when firms use them.
We're looking for contributors in: employment law, M&A, intellectual property, litigation, compliance, real estate, healthcare, immigration, and more.
Common questions
What are prompt containers?
Prompt containers are AI instructions embedded directly in the document. They tell the inference engine how to process specific sections: what tone to use, which clause patterns to follow, what compliance checks to run. When an agent opens the document, it reads these containers and already knows the guidelines without any separate configuration.
What is the .aid format?
How do file type designations work?
Can I use existing documents as .aid files?
See document types in action
Book a 15-minute demo. We'll show you prompt containers, file type designations, and the .aid format with a live document.