docs: improve search_docs ranking for module-system and config queries#129
Merged
Merged
Conversation
Two BM25-ranking fixes (content present, just under-ranked) that take the live search_docs audit from 6/8 to 8/8 (after the docs-fts driver fix in #127): - concepts/modularity.md: lead on the literal phrase "module system" so the query stem matches the title's neighborhood (Q1: rank 5 -> 1). - getting-started/configuration.md: add a "Where Config Defaults Go" heading + lead covering the query terms config/defaults/go, which porter stemming otherwise split from "configuration" (Q6: miss -> rank 1). Verified end-to-end via `claude -p` against the live MCP: F = 8/8, seven at rank 1. Closes the YELLOW item on the search_docs coverage audit (task 000); docs-fts confirmed as the skeleton/devai default. docs-vec remains tracked in #128. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
The final two misses from the live
search_docsaudit (run end-to-end in a real project) were BM25 ranking gaps, not missing content. Two smalldocs-markdownedits take the audit from 6/8 → 8/8:concepts/modularity.md— lead on the literal phrase "module system" so the query stem matches the title's neighborhood (porter stemsmodule→modul, which never matched the titleModularity→modular). Q1: rank 5 → 1.getting-started/configuration.md— add a "Where Config Defaults Go" heading + lead covering the query termsconfig/defaults/go(porter splitsconfigfrom the doc'sconfiguration). Q6: miss → rank 1.Verification
End-to-end via
claude -pagainst the live MCP, and independently re-checked against the real corpus in this repo:Closes out the audit
search_docscoverage audit (task 000) → RESOLVED / GREEN.docs-ftsconfirmed as theskeleton/devaidefault driver.docs-vec(un-buildable on stock PHP) remains tracked in docs-vec: un-buildable / non-functional on standard PHP PDO-SQLite (6 stacked defects) #128.docs-markdown suite green: 17 passed.
🤖 Generated with Claude Code