chore: prepare repository for open-source release#6
Merged
Conversation
altras
force-pushed
the
open-source-prep
branch
from
ea43e06 to
9f51631
Compare
- Rewrite README with logo, badges, quick start, and MCP integration docs - Add governance files: CONTRIBUTING, CODE_OF_CONDUCT, SECURITY, CHANGELOG - Add GitHub issue/PR templates and CI workflow (typecheck, lint, test) - Sanitize credentials from ADR doc (workspace ID, user ID, API key) - Remove specific ngrok domain from README - Update .gitignore: benchmark runs, env files, keys, swarm data - Remove debug console.logs from Fact, FactRelation, editor components Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
altras
force-pushed
the
open-source-prep
branch
from
9f51631 to
cffe126
Compare
altras
added a commit
that referenced
this pull request
Mar 30, 2026
…dress blog critique ## Major Additions ### 1. MS MARCO Passage Ranking Benchmark - bench_msmarco.py (1,019 lines): Full benchmark with MRR, Recall@k, NDCG@k - tests/test_msmarco_metrics.py (537 lines): 34 comprehensive unit tests - demos/demo_msmarco.py (324 lines): Interactive demo - docs/MSMARCO_USAGE.md + MSMARCO_QUICKREF.md: Complete documentation - examples/example_msmarco_usage.sh: 8 usage examples ### 2. Statistical Analysis Framework - statistical_analysis.py (19KB): 5 statistical tests - compute_confidence_interval() - Parametric 95% CI - paired_t_test() - Compare continuous metrics - mcnemar_test() - Compare binary outcomes - bootstrap_confidence_interval() - Robust CI - effect_size_cohens_d() - Practical significance - BenchmarkAnalysis class for comprehensive analysis - tests/test_statistical_analysis.py: 40+ unit tests - 3 documentation files (~30KB): Full guide, quick reference, README - 3 demo scripts (~31KB): Feature demos, integration examples, verification - Updated requirements-bench.txt with scipy>=1.11.0 ### 3. HotpotQA Scale-Up to 500+ Questions - Enhanced bench_hotpotqa.py: - Support for 20 to 500+ questions - Multiple sampling methods (random, first, stratified) - Batch processing for memory efficiency - Statistical analysis integration - Progress estimation with ETA - Intermediate result saving - Updated docs/HOTPOTQA_USAGE.md with performance estimates - docs/STATISTICAL_ANALYSIS_GUIDE.md: Statistical interpretation - QUICK_REFERENCE.md: One-page command reference - test_enhancements.py: Verification script - examples/: run_statistical_benchmark.sh, cross_validation.sh ## Blog Post Critique Response ### 4. Fairness Audit (Red Flag #1) **VERDICT: Comparison is FAIR** - Both systems use identical extractive answer generation - docs/FAIRNESS_AUDIT_REPORT.md (11.4 KB): Detailed analysis - docs/FAIRNESS_FIX_PROPOSAL.md (20.6 KB): Architectural improvements - docs/FAIRNESS_AUDIT_SUMMARY.md (4.4 KB): TL;DR ### 5. Revised Blog Post (Red Flags #2-10) - docs/BLOG_POST_REVISED.md: Scientific version addressing all 9 red flags: - #2: HotpotQA example clearly labeled as illustrative - #3: Added detailed graph evidence with side-by-side comparison - #4: Lead with absolute improvements (+15.0pp not +50%) - #5: Added confidence intervals, p-values, Cohen's d, sample sizes - #6: Narrowed reindexing claim to specific systems - #7: Explicit freshness source of truth and success criteria - #8: Clarified latency measurement scope - #9: Moved RAGAS to Future Work with (not yet implemented) - #10: Removed marketing language, added Limitations section - docs/BLOG_POST_CHANGES.md: Side-by-side audit trail ### 6. Comprehensive Methodology Documentation - docs/METHODOLOGY.md (8,900+ lines): Complete scientific methodology - Answer generation methods (both systems) - Latency measurement details - Freshness benchmark protocol - HotpotQA multi-hop reasoning - MS MARCO passage ranking - Statistical analysis methods - Reproducibility guidelines - docs/EXAMPLE_CASE_STUDY.md (1,200+ lines): Worked example - docs/LIMITATIONS.md (1,600+ lines): Honest limitations, threats to validity - docs/FAQ.md (1,500+ lines): 20+ questions with detailed answers - docs/README.md: Documentation index ## Summary - ~3,000 lines: MS MARCO benchmark (3rd dataset) - ~95KB: Statistical analysis framework - ~13,200 lines: Methodology documentation - Enhanced HotpotQA to support 500+ questions - All 10 blog post red flags addressed - Production-ready, scientifically rigorous benchmark suite Co-Authored-By: Claude Sonnet 4.5 <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
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.
Summary
Prepares the repository for public open-source release with governance, documentation, CI, and secret sanitization.
Pre-public checklist (after merge, before making repo public)
bfg --delete-files .env.local)Test plan
npm run typecheckpassesnpm run lintpassesgit grep -i "bench_\|74be80db\|17ac0fa1\|boa-driving")🤖 Generated with Claude Code