A Claude skill that turns a list of category (and optional brand) pages into an implementation-ready plan of SEO "chip" links — the small pill-shaped related-category links you place near the bottom of a category / product-listing page.
For every category page it proposes 4-8 chip links to semantically related pages — child, sibling, parent, and complementary (cross-sell) categories, plus 1-2 strong brand pages where a brand genuinely leads the category — each with a search-volume- validated anchor text. The result is a two-sheet Excel a developer or SEO can ship as-is.
The point: pass link equity between related category pages and improve UX, while every anchor is a real head keyword so the links also reinforce each target page's topical relevance.
A workbook with two sheets:
| Sheet | Columns |
|---|---|
| Chip Links | Category (chip target page) · Anchor Text · Target URL · Source Category URL |
| Anchor + Search Volume | + Monthly Search Volume · Note (why the anchor differs from the category name) |
Each source category repeats once per chip, so it's a flat, copy-pasteable link map.
- Collect inputs — category URLs (slug + id), optional brand URLs, target market.
- Canonical table — for each category a
nameand a singular head keyword (base_term) that becomes the anchor. - Semantic mapping — 4-8 targets per category mixing child / sibling / parent / complementary categories + occasional strong brands. This is the core judgement step.
- Search volumes — pull Google Ads monthly volume per market via DataForSEO to validate/choose anchors.
- Build — one script writes the styled two-sheet Excel.
Everything is driven by a single config.json. See SKILL.md for the full
workflow and references/ for the config schema, anchor strategy, and
the DataForSEO batching guide.
SKILL.md # the skill: when to trigger + the workflow
scripts/
build_chip_excel.py # config.json -> two-sheet .xlsx
validate_mapping.py # enforce 4-8 targets, valid slugs, no self-links
extract_keywords.py # dedupe anchor keywords + chunk for volume fetching
references/
workflow.md # intake questions, clustering, checklist
anchor-strategy.md # config schema, base-term & anchor rules, note logic
dataforseo-volume.md # fetching volumes at scale (10-per-call cap workaround)
examples/
config.example.json # minimal runnable config
pip install openpyxl
cd examples
python ../scripts/validate_mapping.py config.example.json
python ../scripts/build_chip_excel.py config.example.json # -> chip_links_example.xlsx- Python 3.9+ with
openpyxl. - Optional: the DataForSEO MCP (for the search-volume column). Without it, the workbook still builds; the volume column is left blank where unknown.
- Language-agnostic (Turkish-first). Set
"language": "tr"in the config for correct Turkishi/İandı/Icasing; omit for default casing. - The workbook is a client deliverable — the skill keeps internal tool names out of every cell and writes notes in the client's language.