The old has_similar_post() used WordPress search ('s' parameter) which
matches ANY word in the query. On a niche peptide blog, every topic
shares vocabulary like "BPC-157", "peptide", "dosage" — so all 125
ideas were false-positive deduped against existing posts, producing
zero viable article ideas.
New approach: extract meaningful keywords (strip stopwords), compare
against each existing post title, and only dedup when ≥60% of the
topic's keywords overlap with a single title. Also caches the title
list per pipeline run to avoid repeated queries.
Added summary logging so we can see exactly how many ideas pass
through each filter stage.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- ContentAnalyzerTest: pass Cost_Tracker as 2nd constructor arg
- IdeaScorerTest: stub get_option for topic_exclusions
- CostTrackerTest: stub get_option for monthly_budget_usd
- OpenRouterPricingTest: stub get_transient with cached model data
- OpenRouterProviderTest: stub get_option for API key + get_transient
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
All 14 test files updated to use correct constructor patterns (array-based
value objects, dependency injection for ContentAnalyzer, etc.) and verified
method names against source.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
