Marketing-accuracy pass: real star count, honest numbers, full hedge, semantic promoted - #41
Merged
Merged
Conversation
…hedge From a three-way claims audit against the committed benchmark reports: - Nav star count is now fetched from the GitHub API at build time (the hardcoded '1.2k' default rendered site-wide; the repo has 22 stars) and hides when the API is unreachable. - Corrected figures: obsidian-mcp-rs 6.9→6.1 ms @1k; Pillars token figure 740 (fs adapter) → 507 (seekstone's own number); the 47,000× on /benchmarks re-attributed to REST-proxy servers (the plugin alone is ~31,000×); '1–6 ms' range → '6.2 ms mean (1.2–18.7 across queries)'; 'peaked at 370.9 MB' → averaged; latency-range note now spans the real 430–2,714 ms with the obsidian-mcp-rs exception; tiers copy no longer claims 'over a second' for servers topping out at 958 ms. - Counts and methodology: 'six servers' → eight in three places (incl. FAQPage JSON-LD); methodology field list gains obsidian-mcp-rs and obsidian-tc; 'median (p50)' → mean of per-query medians; '18 tools' → 19 on /docs/getting-started. - Tool cards: replace_in_note no longer claims first-occurrence-only (it replaces all by default), patch_note describes its real three operations, delete_note's permanent wording clarified. - Full hedge pass: title/OG/footer → 'fastest in open benchmarks'; hero chip → '440× faster than any server we tested'; 'works offline' / 'zero network calls' scoped to runtime everywhere incl. JSON-LD; 'published numbers' headline gains 'as far as we know'; competitors.ts write-safety answer updated (the cross-server safety matrix IS committed now); stale competitor-set copy fixed. - Semantic search promoted from tradeoff-only to a measured feature: new /benchmarks section with the committed golden-set numbers (70% vs 30% hit@5, ~14 ms, 100% exact-title retention) and an explicit single-vault / no-competitor-comparison caveat; setup instructions now include the mandatory SEEKSTONE_SEMANTIC=1; tradeoff card states the real costs (background embed, vector cache); hero and Tools copy own both latency numbers; getting-started covers enablement. - version.ts fallback bumped 0.7.2 → 0.15.0. Build verified.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This branch was successfully deployed
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
1 participant
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.
Companion to seekstone's marketing-accuracy PR — see the commit message for the itemized findings/fixes from the three-way claims audit. Every number now traces to a committed baseline report; every superlative carries its scope; semantic search is promoted with our own measured numbers and an explicit no-competitor-comparison caveat. Opened for review — do not merge until the seekstone repo PR is reviewed alongside it.