B2BITS is introducing a new generation of product enablement assets designed for the era of AI-assisted software development.

As software development moves toward AI-assisted and agentic workflows, traditional API documentation alone is no longer sufficient. AI tools require structured, reliable, and context-rich product knowledge to generate accurate code, automate configuration, and support troubleshooting.

To address this shift, FIX Antenna, FIXEdge, and related B2BITS products will include AI-ready enablement packages with Agent Skills optimized for both human developers and AI agents.

These packages are designed to:

  • Shorten client onboarding cycles
  • Improve the correctness of AI-generated integration code
  • Reduce avoidable support requests
  • Accelerate product adoption and time to value

Agent Skills: AI Knowledge Packages

Each supported B2BITS product distribution will include an Agent Skill: a structured, version-pinned knowledge package optimized for AI-assisted development, configuration, and troubleshooting.

An Agent Skill complements standard product documentation by organizing product knowledge in a form that AI assistants can reliably use. It helps AI tools understand not only individual APIs and configuration options, but also how those elements should be used together in real integration workflows.

Agent Skills may include:

  • API and configuration usage guidance
  • Recommended integration patterns
  • Session lifecycle rules
  • Configuration models and constraints
  • Business workflow examples
  • Troubleshooting guidance
  • Common failure modes
  • Known anti-patterns and misuse cases
  • Version-specific behavior and compatibility notes
  • References to approved examples and implementation patterns

AI-Ready FIX Antenna Libraries

FIX Antenna libraries for C++, Java, .NET, Python, and Rust will be delivered with additional AI-ready enablement artifacts.

These artifacts allow AI assistants such as coding agents to understand not only APIs but also correct usage patterns and product behavior.

Each FIX Antenna library will include a product-level AI onboarding guide describing:

  • Library purpose and architecture
  • Common integration scenarios
  • Session lifecycle
  • Initialization and shutdown procedures
  • Threading model
  • Error handling approach
  • Memory ownership rules
  • Performance considerations

This guide will serve as a structured reference for AI agents and developers working with FIX Antenna libraries.

Enhanced API documentation

Public APIs will be enriched with additional documentation, including:

  • Richer inline comments
  • Parameter descriptions
  • Lifecycle expectations
  • Error conditions
  • Usage constraints

The goal is to make API behavior clearer and more predictable for both human developers and AI-assisted coding tools.

Integration examples

FIX Antenna libraries will include executable reference examples and integration patterns that demonstrate correct API usage, lifecycle management, and common FIX protocol workflows.

The examples will cover two levels of guidance:

Examples

Core Technical Patterns

Examples demonstrating correct use of the library for common technical tasks, including:

  • Initiator session setup and lifecycle management
  • Acceptor session setup and lifecycle management
  • Logon, logout, reconnect, and shutdown handling
  • Sequence number management
  • Resend request and recovery handling
  • Custom message parsing and processing
  • Error handling, logging, and diagnostics
  • Configuration loading and validation

Scenario-Based FIX Workflows

Reference workflows demonstrating how typical FIX message flows can be implemented using FIX Antenna, including:

  • Market data subscription flow
  • New order, cancel, and cancel/replace flow
  • Execution report processing
  • Drop-copy message consumption
  • Session recovery after disconnecting
  • Handling of custom or venue-specific FIX fields

Each example will clearly document its scope, assumptions, and required adaptation points, including what must be customized for a specific venue, counterparty, asset class, or production environment.

Example Business Requests

  • "Create a FIX Antenna C++ application connecting to an exchange using FIX 4.4 with automatic reconnect."

AI can generate correct starter code based on approved examples.

AI-Ready FIXEdge C++ and FIXEdge Java

FIXEdge C++ and FIXEdge Java will also include AI-ready artifacts that support AI-assisted configuration, message routing, transformation, validation, and business logic development.

The goal is to help clients move from business-level requirements to implementation-ready FIXEdge configuration more efficiently. Instead of starting from blank configuration files or low-level documentation, developers and support teams will be able to use AI assistants to generate draft configurations, explain existing setups, suggest modifications, and identify the areas that require venue- or counterparty-specific customization.

These AI-ready artifacts will help AI tools understand FIXEdge concepts such as:

  • Session configuration
  • Message routing
  • Message enrichment
  • Transformation rules
  • Validation rules
  • Business rule processing
  • Error handling and reject logic
  • Deployment-specific configuration boundaries

AI-generated outputs would remain subject to developer review, testing, certification, and client-specific operational controls.

Example Business Requests

With these artifacts, AI assistants can help translate business-level requests into FIXEdge configuration drafts or implementation guidance, such as:

  • “Route all NewOrderSingle messages from Client A to Broker B and enrich them with the required custom tags.”
  • “Transform incoming client FIX messages into the exchange-specific message format required by Venue X.”
  • “Create validation rules for incoming orders, including required fields, allowed values, and reject behavior.”
  • “Add routing logic that sends drop-copy messages to a monitoring session while preserving the original execution flow.”
  • “Explain how this existing FIXEdge configuration handles order routing, message transformation, and error scenarios.”
  • “Generate a draft configuration for a new client session based on an existing counterparty setup, highlighting all fields that must be reviewed or customized.”
  • "Create a rule that receives FIX orders from MQ, validates them, enriches fields, and sends them to a FIX session."

Business Layer AI Enablement

Building on the AI-ready FIXEdge artifacts, B2BITS will provide structured guidance and reference patterns to support AI-assisted development of FIXEdge Business Layer assets.

These artifacts will help AI tools translate business-level requirements into implementation-ready drafts, configuration fragments, scripts, transformation logic, and extension scaffolding. The goal is to accelerate development while keeping the final implementation under developer review, testing, certification, and client-specific operational control.

FIXEdge C++ Business Layer

For FIXEdge C++, AI enablement artifacts will support the creation and maintenance of:

  • XML, JavaScript-based configuration fragments and templates
  • XSLT transformation templates
  • Routing rules and session-selection logic
  • Message transformations, enrichment and normalization logic
  • Validation and reject-handling patterns
  • Scaffolding for custom C++ Business Layer extensions

These artifacts will help AI assistants understand how FIXEdge C++ configuration, transformation, and extension mechanisms should be used together to implement client-specific workflows.

FIXEdge Java Business Layer

For FIXEdge Java, AI enablement artifacts will support the creation and maintenance of:

  • Groovy script templates
  • Java business logic component scaffolding
  • Message transformation logic
  • Validation rules
  • Workflow extensions // what are these? Why not just same routing/validation/transformation as in C++?
  • Routing and enrichment logic
  • Error-handling and reject-processing patterns

These artifacts will help AI assistants generate consistent draft implementations that align with FIXEdge Java Business Layer concepts and are easier for developers to review, adapt, and test.

Other AI-Assisted Client Use Cases

Troubleshooting and Diagnostics

Agent Skills can also support troubleshooting by helping AI assistants interpret product behavior using provided logs, configuration, sequence numbers, and session settings.

Example request: “Why is my FIX session continuously sending ResendRequest messages?”

Based on the available diagnostic information, the assistant can help identify likely causes, such as:

  • Sequence number mismatch
  • Missing or unavailable persisted messages
  • Incorrect recovery settings
  • Session reset behavior
  • Counterparties resend behavior
  • Misaligned FIX session configuration

The assistant can then suggest investigation steps and possible remediation options based on documented product behavior and known failure modes.

Benefits for Clients

Benefits

Faster Implementation

Agent Skills help clients move faster from product documentation and business requirements to working integration drafts, configuration fragments, and implementation patterns.

For common integration and configuration tasks, this can reduce the time required to create first working versions and accelerate the path to client-specific implementation.

Better Implementation Consistency

AI-generated code and configuration drafts can be aligned with official B2BITS examples, documented product behavior, and recommended implementation patterns.

This helps reduce variation between implementations and lowers the risk of common integration mistakes.

Lower Operational Risk

Agent Skills package knowledge about lifecycle behavior, recovery scenarios, threading expectations, configuration constraints, and known failure modes.

This helps AI assistants generate outputs that are more aware of operational behavior, especially in areas such as reconnect handling, resend recovery, session management, and error processing.

AI assistants can use structured product guidance covering recovery, threading, and session behavior.

Reduced Dependency on Expert Knowledge

Agent Skills make product knowledge more accessible to developers, implementation teams, and support engineers.

New team members can become productive faster by using AI assistants to explain APIs, configuration models, integration patterns, troubleshooting steps, and existing deployments.

Easier Maintenance and Upgrades

Agent Skills can also help clients maintain, extend, troubleshoot, and upgrade existing FIX Antenna and FIXEdge deployments.

AI assistants can help explain existing code or configuration, identify areas affected by product version changes, suggest migration steps, and highlight parts that require expert review.

Conclusion

AI-ready artifacts transform B2BITS products from traditional software libraries into agent-friendly platforms.

FIX Antenna and FIXEdge will not only provide APIs and documentation but also the knowledge required for AI assistants to correctly build, configure, troubleshoot, and extend production trading systems.

Clients can also maintain, extend, troubleshoot, and upgrade existing FIX Antenna and FIXEdge deployments faster, with less dependence on product experts and support tickets