Repository navigation
Product analysis (PRD, PM, UX/UI, algorithm) + landing conversion edits - #31
Merged
Merged
Conversation
…rsion edits - docs/product/PRD.md: product requirements doc - docs/product/PM-FEATURE-ANALYSIS.md: prioritized feature opportunities - docs/product/UX-UI-AUDIT.md: UX/UI findings + quick wins - docs/product/EVALUATION-ALGORITHM-ANALYSIS.md: Trust Score / SkillForge analysis + hardening plan - Landing: benefit-first hero copy + ease cues, fix count inconsistency (4,200+ → hundreds), unify Trust Score naming, correct license label (Apache 2.0 / CC BY-SA)
Landing page: - Remove SocialProof entirely (illustrative testimonials erode trust) - Remove redundant SkillLayerManifesto + NetworkSection; ease-first reorder - MCP tool picker (Claude/Cursor/ChatGPT/Other) swaps connect steps - Single primary CTA + quiet secondary link; benefit cues in hero - Unify counts + Trust Score naming; fix license label Trust Score v2 (evidence-gated, reproducible): - src/lib/trust/scoring.ts: Wilson lower-bound, confidence gating with verified flag (no more 0.5 defaults), freshness term, multi-dim vector - supabase migration recompute_trust_scores_v2() mirrors it in SQL + repoints cron - src/lib/adversarial/holdout.ts: deterministic salt-rotatable train/holdout split - tests: trust-scoring + adversarial-holdout (12 cases, all green) - docs updated to reflect what shipped
criptogus
marked this pull request as ready for review
May 29, 2026 18:40
This was referenced May 29, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What this delivers
A full product analysis of Super Agent Skill plus high-confidence conversion edits to the landing page. Analysis lives in
docs/product/.1. PRD —
docs/product/PRD.mdProblem, vision, personas, the four primitives, current flows/pricing, the Trust Score formula, proposed North-Star metrics, instrumentation, risks, roadmap.
2. Landing page conversion edits —
src/routes/index.tsxGoal: simpler message, lead with ease + benefits, fix credibility bugs.
3. PM feature analysis —
docs/product/PM-FEATURE-ANALYSIS.md15 prioritized opportunities (RICE-framed) across activation, trust moat, flywheel, ARPU, virality — with a suggested sequencing and the metric each bet should move.
4. UX/UI audit —
docs/product/UX-UI-AUDIT.mdCritical (count/naming/license/fake-testimonial risks), high-impact conversion (message overload, jargon above the fold, competing CTAs, MCP literacy), plus medium/polish and app-surface notes. The quick wins are implemented in this PR.
5. Evaluation-algorithm analysis —
docs/product/EVALUATION-ALGORITHM-ANALYSIS.mdEnd-to-end breakdown of the proprietary stack (adversarial harness, type-aware evaluator weights, the daily Trust Score formula, SkillForge evolution loop), its weaknesses (substring grading is gameable, 0.5 defaults reward absence of evidence, no recency/confidence weighting, static suite, Goodhart risk in the Forge gate), and a Trust Score v2 hardening plan (Wilson lower-bound, evidence-gating, recency/model-aware scores, LLM-judge+holdout suite, reproducible signed methodology) to make the moat uncopyable.
Notes
node_modulesisn't installed (vite/client and @eslint/js missing). The edits are JSX copy/list changes and were verified by inspection.https://claude.ai/code/session_01BnYrQNoKNNxPCSmV2suWna
Generated by Claude Code