analyze-keyword-opportunities

name	analyze-keyword-opportunities
description	Recommend new posts to add to the content backlog and existing posts to refresh, based on GSC + DataForSEO + posts manifest. Caps suggestions to keep signal high.

analyze-keyword-opportunities

Reads posts.json, search-console.json, dataforseo.json. Emits insights.opportunities.

How to invoke

python .claude/skills/analyze-keyword-opportunities/scripts/analyze.py [--week 2026-W18]

Hybrid skill: Python identifies and scores candidate keywords; Claude writes the rationale, suggested_angle, and recommended_changes prose for the top picks.

Output

{
  "new_post_suggestions": [ /* up to 3 */ ],
  "post_refresh_suggestions": [ /* up to 3 */ ]
}

Schema per plan.md §3.9. Both arrays include a rationale field — the prose is the whole point. Without strong rationale, suggestions feel like spam.

Scoring

• New post candidates: discovered keywords (impressions ≥100 over 90 days, your_position 11–40) where no existing post declares this keyword as primary or secondary. Score by volume × (1 / difficulty) × (1 / position).
• Refresh candidates: existing posts with declining position trend AND a striking-distance keyword they could plausibly capture with updates.

Notes

• 3 + 3 cap is intentional. More than that becomes noise and you stop reading.
• Internal linking suggestions for new posts come from posts.json — surface 2-3 high-authority existing posts that semantically relate to the new keyword.