2 commits

Author	SHA1	Message	Date
Dhravya	771be5cef8	fix: apply review feedback — fix double data/ prefix, semaphore bug, resume bug, consolidate duplicated code ... - Fix worker.py writing to data/data/ instead of data/ (critical path bug) - Fix semaphore recreation on every call due to checking _value instead of capacity - Fix questions.py resume returning raw string instead of list[dict] - Fix prompts/file_gen.py reading 'summary' instead of 'brief' from manifest - Extract shared unwrap_json_list() and truncate_to_tokens() into utils.py - Remove redundant validation report writes in generate.py - Remove unused imports and dependencies - Fix f-string logger calls to use lazy %s formatting - Move calendar import to top-level in validator.py - Use write_text() for atomic writes in repair_files() - Strengthen test_resume_support to assert return type	2026-04-28 23:49:23 +00:00
Dhravya	cba994be3f	feat: add eval corpus data generator ... 7-phase pipeline for generating synthetic multi-file corpora: 1. Scenario Brief (SCENARIO.md) - world-building 2. Fact Registry (facts.json) - consistency source of truth 3. File Manifest (manifest.json) - per-file briefs 4. Clustering - topological sort + fact registry sharding 5. Parallel File Generation - concurrent workers 6. Validation - cross-reference & consistency audit 7. Question Generation - 10 eval questions per corpus Supports all 13 data points (dp_001 through dp_013, 5 to 10,000 files). Uses Gemini 2.5 Pro by default. Includes resume support, validation, and 219 unit tests.	2026-04-28 23:24:42 +00:00