test(01-03): enforce the transport seam statically and in a real browser

feat(01-03): add both transports, the standalone stage page and the Gradio component

docs(01-03): restore the crash-recovery position clobbered by begin-phase

docs(01-03): record the crash-interrupted state of the transport seam plan

feat(01-03): add the single facade builder and the shared turn loop

feat(01-03): add the transport-agnostic VRM stage, viseme player and audio queue

docs(01-01): complete hosting and VRM character gates plan

feat(01-01): add the VRM tutor character and record its license facts

docs(01-01): record the ZeroGPU hosting decision and wire the Space remote

docs(01-04): complete Japanese voice plan

docs(01-04): record the measured cold-start cost of the lazy runtime fetch

test(01-04): capture ground-truth synthesis fixtures for AVTR-02

feat(01-04): implement synthesize() and the shared voice dataclasses

feat(01-04): add VOICEVOX runtime assets, requirements.txt and setup docs

docs(01-02): complete local foundations plan

test(01-02): add deterministic silence and cafe-noise audio fixtures

test(01-02): add shared fixtures and --space-url plumbing

chore(01-02): bootstrap Python 3.12 toolchain, Git LFS and package skeleton

docs(01): complete files_modified frontmatter on plans 05, 07, 09

fix(01): revise plans per checker feedback

docs(01): create phase plan (10 plans, 6 waves)

docs(01): add validation strategy

docs(01): research phase 1 voice + avatar loop skeleton

docs: create roadmap (6 phases)

docs: define v1 requirements

docs: project research

chore: add project config

docs: initialize project