- Feature
- Diagnose AL MCP server activity with file logging
- Availability
- Generally available (October 2026)
- Industry focus
- Insurance brokerage and managing general agents
Insurance brokers operate under close regulatory supervision, and their technology teams increasingly use AI coding agents to build and maintain Business Central extensions. Commission calculations, client money reconciliations and insurer settlement files all depend on that code. When AI is involved in changing it, someone needs to be able to see what happened. Business Central 29.0 adds file logging to the AL MCP server for exactly that reason.
The industry challenge
A mid-sized broker typically extends Business Central for:
- Commission and fee calculations by insurer and product
- Client money and insurer money segregation
- Settlement statements and bordereaux
- Integration with a policy administration system
When developers use AI coding agents, the agents connect through the AL MCP server to read projects, analyze code and propose changes. Two practical problems follow:
- Troubleshooting. When the agent tooling fails intermittently, such as a tool call timing out or a project not loading, there's often no trace to investigate.
- Oversight. Compliance and IT governance teams want to understand what agents did during development, especially for code that touches client money.
Without logs, both questions get vague answers.
What Microsoft delivered in Business Central 29.0
According to Microsoft, the AL MCP server can now:
- Write diagnostics to a configurable file
- Use a level of detail suited to the troubleshooting at hand
- Turn file logging off when it isn't needed
Microsoft says this lets developers and administrators keep AL MCP server diagnostics for troubleshooting intermittent failures and for reviewing agent activity. Adjustable detail means teams can capture rich information during an investigation without producing excessive logs all the time.
Where the value shows up
| Business area | Impact for an insurance broker |
|---|---|
| Reliability | Evidence to fix intermittent failures in AI development tooling |
| Governance | A record of agent activity during development work |
| Audit readiness | Supporting material when reviewing changes to sensitive code |
| Cost control | Detailed logging only when needed, not permanently |
| Partner collaboration | Logs that can be shared with a Business Central partner |
This isn't a business process log. It's a development tooling log. But in a regulated firm, being able to show how AI-assisted changes were made is part of a credible change-management story.
A worked scenario
Consider a commercial lines broker with an in-house developer and a Business Central partner (an illustrative example). The developer uses an AI coding agent to help refactor the commission calculation extension ahead of a new insurer agreement.
Halfway through, the agent's tool calls start failing at random. With file logging switched on at a detailed level, the developer captures a log showing that a connection setting is wrong for one environment. The partner fixes it in minutes.
Later, during the quarterly IT governance review, the compliance officer asks how the AI agent was used on the commission code. The developer shows the retained logs alongside the normal source control history and code review records. Then they reduce the logging level back to normal.
Adoption considerations
- Set a default level. Use a low level of detail day to day and raise it during investigations.
- Decide on retention. Agree how long logs are kept, in line with your IT and data policies.
- Protect the logs. Diagnostics may include project details or environment names. Store them securely.
- Combine with source control. Logs show agent activity. Pull requests and reviews show what was accepted.
- Include it in your AI policy. Make logging part of your rules for using AI in development.
The bottom line
Insurance brokers can benefit from AI-assisted development, but only if they can explain how it was used. AL MCP file logging in Business Central 29.0 gives development teams a practical way to troubleshoot their AI tooling and keep a record of agent activity. It's a small feature that supports a big principle: AI in regulated businesses should be visible, not invisible.
If your firm is starting to use AI coding agents on Business Central, it's worth setting logging and governance rules with an experienced adviser from day one.
Source: Microsoft roadmap 573349