From 41ce85548de0729f1a882895f44aec98ec9609f4 Mon Sep 17 00:00:00 2001 From: David de Hilster Date: Wed, 15 Jul 2026 11:58:33 -0400 Subject: [PATCH] docs(help): note the NLPPlus Interface analyzer template Add a shortcut callout to the package usage page: when creating a new analyzer, choosing the "NLPPlus Interface" template gives an analyzer pre-wired for the two-way package exchange -- the json2kbb pass is already in the sequence before the tokenizer and an output pass already writes output.json -- so users can skip the manual json2kbb setup. Co-Authored-By: Claude Opus 4.8 (1M context) --- Help/markdown/vscode/packages.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/Help/markdown/vscode/packages.md b/Help/markdown/vscode/packages.md index 1601d95..9f87631 100644 --- a/Help/markdown/vscode/packages.md +++ b/Help/markdown/vscode/packages.md @@ -119,8 +119,19 @@ knowledge base. into a `.kbb` knowledge‑base file the engine loads on the run. If it isn't in the sequence (before the tokenizer), the JSON is written but **never loaded into the KB**. +> **Shortcut — start from the "NLPPlus Interface" template.** When you create a new +> analyzer (Analyzers view → **New Analyzer**), pick the **NLPPlus Interface** template. +> It comes pre‑wired for exactly this two‑way exchange: the `json2kbb` pass is already in +> the sequence **before the tokenizer**, and an `output` pass already writes `output.json` +> for the package to read back. You get a working import/export analyzer out of the box — +> just replace the sample `kb/user/data.json` and add your own rules. Choose this template +> and you can skip the manual "add the `json2kbb` pass" step below. + ### Step 1 — add the `json2kbb` pass to the analyzer (one time) +*(Skip this step if you started from the **NLPPlus Interface** template — it already has +the `json2kbb` pass and the `output.json` export set up.)* + In the VS Code NLP++ extension, open the **Sequence** view and **right‑click the tokenizer** (or any pass) → **Insert Python Library Pass (Before Tokenizer)** → choose **`json2kbb`**. It must run **before the tokenizer** so the data is in the KB before