Agent Workflow
How ProofKit closes the loop for coding agents building FileMaker apps.
ProofKit is designed for agents that can do more than produce snippets. The workflow gives the agent real FileMaker context, a project it can edit, and feedback it can use to fix mistakes.
The loop
What each step needs
- Next change: the user asks for a new screen, behavior, or refinement.
- Read FileMaker schema: the MCP server exposes file metadata and data context to the agent.
- Write app code: the agent edits a normal web project using React, TypeScript, Tailwind, and shadcn/ui.
- Run checks: TypeGen and TypeScript catch many schema and field-name mistakes before runtime.
- Preview in browser: the agent can inspect the app as it runs, not just reason about code.
- Fix issues: errors, screenshots, and build output become feedback for the next edit.
- Deploy to FileMaker: the final bundle is pushed into the file and loaded by the Web Viewer.
Why this matters
Without the loop, you are usually copying FileMaker metadata into a chat, pasting code back into a project, discovering errors manually, and asking the model to try again. ProofKit gives the agent enough context and feedback to handle more of that cycle directly.
Boundaries
ProofKit v2 focuses on agentic coding of Web Viewer apps. It does not make the agent a full FileMaker schema editor. Scripts, tables, fields, layouts, and value lists remain outside the supported writing scope for this release.