post: How We Delivered 70 Pages of Technical Docs with AI

[GBurgardt]

Note

Low Risk
Low risk: adds a new unpublished Markdown post only; no runtime/application code changes beyond potential missing/incorrect referenced assets or frontmatter expectations.

Overview
Adds a new draft blog post how-we-delivered-70-pages-of-technical-documentation-using-ai.md describing an iterative workflow for producing a large technical implementation document using debate, real-time transcription, LLM drafting, web-search verification, and coding-agent proofs of concept.

Includes frontmatter metadata (title/description/tags/cover image) and multiple referenced images in ./assets to illustrate the process and a concrete attachment-persistence finding.

^{Written by Cursor Bugbot for commit 331cdfa. This will update automatically on new commits. Configure here.}

[github-actions]

Dry-run output:

[UPDATED] [PUBLISHED] Code Faster in Cursor: A Pragmatic Guide to Voice Pr… 
[UPDATED] [PUBLISHED] How AI Reflects Your Thinking                         
[CREATED] [DRAFT]     How We Delivered a Full Technical Implementation Doc… 
[UPDATED] [PUBLISHED] Automate Any Repetitive Task with MCP                 
[UPDATED] [PUBLISHED] Forget the Hype: Agents are Loops                     
[UPDATED] [PUBLISHED] WhatsApp + MCP: automatic audio transcription