feat(ai-partner): per-chapter chip pool generator (#1461) - #1494
Merged
Conversation
Build-time pipeline that pre-generates the FAB peek chip pool so runtime chip loading is ~$0. - _tools/build_prompts.py — orchestrator. Dry-run + incremental + entity-only + chapter-only modes, SIGINT checkpoint, Haiku call with retry + bounded label validation (6-10 words). - _tools/build_prompts_variants.py — 6 fixed profile variants (generic balanced, Reformed x2, Jewish x2, Catholic gospels) with tradition + scholar + genre bias and a seed instruction string per variant. - _tools/build_prompts_entity.py — templated chips for people/places/ debates (993 entities; zero LLM cost). - _tools/build_sqlite_schema.py — new precached_prompts table + (entity_type, entity_id) index. - _tools/build_sqlite_loaders.py — populate_precached_prompts(conn) that merges prompts.db → scripture.db; graceful no-op when absent. - _tools/build_sqlite.py — wired in after embeddings. - _tools/validate_sqlite.py — new section 7 PRECACHED PROMPTS with warn-don't-fail for missing chapter rows (CI lacks the Anthropic key). - _tools/content_writer.py — save_chapter() now flags dirty chapters in _tools/prompts_manifest.json alongside the embeddings manifest, so --incremental regens only touched chapters. Entity chip generation verified locally (993 rows for 313 people + 372 places + 308 debate topics). Dry-run estimates ~\$7 one-time for full chapter build (1,189 chapters × 6 variants). 15 unit tests cover: profile variants, entity chip shape, chip validation (word count, source types, empty seed), context assembly. Full DB rebuild + validate still green with graceful skip-with-warning when prompts.db is absent. https://claude.ai/code/session_01Pht3kzgdvkn81DDfL9SnFe
Content Pipeline Results✅ All pipeline checks passed
|
Test Results✅ All tests passed
Coverage
⏱️ Duration: 76.2s |
Uh oh!
There was an error while loading. Please reload this page.
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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Closes#1461. Phase 3 of epic #1446.
Summary
Pre-generates the FAB-peek chip pool at content build time so runtime cost is ~$0.
prompts.db, merged intoscripture.dbby the existing build orchestrator.New files
_tools/build_prompts.py— orchestrator with--dry-run,--incremental,--entity-only,--chapter-only, cost prompt >$1, SIGINT checkpoint._tools/build_prompts_variants.py—PROFILE_VARIANTSlist._tools/build_prompts_entity.py— people/place/debate chip templates + meta iterators._tools/test_build_prompts.py— 15 unit tests.Modified files
_tools/build_sqlite_schema.py—precached_promptstable._tools/build_sqlite_loaders.py—populate_precached_prompts()loader with graceful skip._tools/build_sqlite.py— wired in._tools/validate_sqlite.py— new section 7 PRECACHED PROMPTS._tools/content_writer.py—save_chapterflags the prompts manifest alongside embeddings..gitignore—prompts.db,prompts_manifest.json.Verification
python3 _tools/build_prompts.py --entity-only— writes 993 entity rows in seconds.python3 _tools/build_prompts.py --dry-run— estimates ~$7 one-time for full chapter build (1,189 × 6 variants).python3 _tools/build_sqlite.py && python3 _tools/validate_sqlite.py— green with warn-don't-fail when prompts.db is absent.python3 _tools/test_build_prompts.py— 15/15 pass.Out of scope
https://claude.ai/code/session_01Pht3kzgdvkn81DDfL9SnFe