indexgap init

Read the project and record what makes it different.

Works out where the pages are, the site URL, the content type and the dataset, writes indexgap.json, installs the agent skills into .claude/skills/ and adds the working files to .gitignore. Run once per project.

Usage

usage: indexgap init [-h] [--site SITE] [--content CONTENT]
                     [--profile PROFILE] [--dataset DATASET] [--key]
                     [--agents] [--force] [--lang {en,ru}]
                     [dir]

positional arguments:
  dir                project directory

options:
  -h, --help         show this help message and exit
  --site SITE        the site address, if it could not be worked out
  --content CONTENT  directory of pages
  --profile PROFILE  content type: catalog, events, ugc, product
  --dataset DATASET  keyword CSV
  --key              mint a new IndexNow key for this project
  --agents           create AGENTS.md for Codex if it is missing
  --force            overwrite an existing indexgap.json
  --lang {en,ru}     output language: en or ru. By default from INDEXGAP_LANG
                     or the system locale, otherwise English

The rest of the pipeline

• planAudit the dataset before a single page is generated
• checkThe full audit of generated pages
• briefTurn findings into work orders
• sitemapBuild a sitemap from what actually exists
• notifyPush new and changed URLs to IndexNow
• portfolioOne table across every project
• profilesShow the built-in content profiles
• doctorThe crawl-to-index funnel
• citeMeasure whether AI search cites you

plan →