Check and fix Japanese PR TIMES press releases with rule tables and OpenAI

Quick overview

This workflow collects a Japanese press release draft via an n8n form, checks it against rules stored in an n8n Data Table, uses OpenAI to revise the text to match those rules, then runs a second OpenAI review and shows the findings and corrected draft on a results page.

How it works

1. Receives a press release title, subtitle, and body from an n8n form submission.
2. Loads all checks from the pr_release_rules n8n Data Table and compiles them into a single rules list.
3. Evaluates the draft against the rules (length limits and keyword-based flags) and produces a findings list with per-field details and advice.
4. Sends the rules, findings, and original draft to OpenAI to generate a corrected title, subtitle, and body plus a list of changes.
5. Sends the corrected draft to OpenAI again to verify it still follows the rules and returns a pass/fail result with any remaining issues and a reviewer note.
6. Displays the original findings, the corrected draft, the change list, and the second-review results in a completion page.

Setup

1. Add an OpenAI credential and select a chat model in the OpenAI Chat Model node.
2. Create an n8n Data Table named pr_release_rules with columns rule, target, check_type, value, and advice, then populate it (optionally by running the sample-rules branch once).
3. Open the form trigger URL, test-submit a draft, and adjust your rules (for example subtitle max length and disallowed wording) before enabling the workflow.