Feat: capture inbox (Phase B) — phone captures merge into the library

Feat: /capture endpoint core (Phase A) — single-post extraction

Feat: living-library merge polish (Phase 3)

Feat: living-library cost-skip on re-extract (Phase 2)

Feat: living-library merge on import (Phase 1)

Feat: client-side library persistence (localStorage restore card)

Test: export.build_kml coordinate invariants (regression for coord-less KML)

Feat: multi-venue extraction for transcription path + tests

Fix: URL fetch works on hosted via Instagram's embed/captioned endpoint

Test: add pipeline tests for multi-venue extraction and URL fetch helpers

Fix KML round-trip: embed City/State in description CDATA; read them back on import

UX: drop Exports tab, auto-spin Roulette on first visit

Tests: add test_pipeline.py + test_routes.py; consolidate TASKS into TODO

Web UI: tab-bar SPA, KML/CSV import, refactor index.html

Fix: KML invalid-XML chars rejected by Google My Maps

Fix: chatbot prompt no longer hardcodes a (wrong) post count

Add end-to-end pipeline test through the real routes (CI)

Fix: chatbot get_prompt KeyError (500 on Prepare extraction file)

#19: public-URL hardening

#18: extend guardrail test for client-geocode + set-coords

Demarcate the two deployment modes + add guardrails, CLAUDE.md

Benchmark: regenerate oracle with is_place labels (fair model comparison)

Benchmark: restore analyze_post, run Opus, persist results as CI baseline

Chore: remove temp task-planning files from tests/

Feature: file upload validation + keyword pre-filter before LLM calls

Feature: in-app results browser (R2–R4 backend + R6–R11 frontend)

Results browser #R5 + #R1: geocode_one helper + GET /results endpoint

Feature: quality benchmark suite + CI regression test