Best Approach for Generating User/Support-Focused Docs from Legacy Code?

Hi everyone,

We’re exploring how to generate technical docs from a legacy codebase — not for developers, but for support engineers and users. The idea is to capture not just functions, but also cross-module workflows and system behavior, so teams can troubleshoot and understand the software more easily.

Rough phases:

  1. Parse codebase → extract structure & relationships

  2. Build context → let AI “walk” across files/modules

  3. Generate docs → practical Markdown docs (support/user oriented)

  4. Publish → ideally in Confluence

Looking for community input on:

  • Parsing tools you’ve found effective

  • Best practices for AI cross-file context

  • Strategies for docs that are support/user friendly

  • Tips for making the pipeline sustainable across releases

We’re determined to move fast and minimize early missteps — any lessons or tools you can share would be invaluable.

Thanks!

Thank you https://community.openai.com/t/best-approach-for-generating-user-support-focused-docs-from-legacy-code/1350219?u=balaji_dilip_singh_r