Add placeholder variable support in timeline JSON for dynamic rendering

[Copilot]

Many workflows generate timeline JSON programmatically and need to inject runtime values (paths, titles, timestamps) without manual edits. This adds ${VAR_NAME} placeholder resolution to render_video, enabling reusable JSON templates.

Changes

• pavo/pavo.py — new resolve_placeholders(data, variables) helper recursively substitutes ${VAR_NAME} tokens in any string value of a JSON-compatible structure; raises ValueError naming any unresolved placeholder. render_video gains an optional variables=None kwarg; substitution runs after json.load() and before schema validation.
• tests/test_render_video.py — unit tests for resolve_placeholders (single/multiple substitutions, nested dicts/lists, non-string pass-through, missing-variable error) and integration tests for render_video with variables=.
• README.md — new "Placeholder variables (dynamic pipelines)" section with template JSON and usage example.

Usage

# template.json contains "${ASSET_DIR}/intro.mp4", "${ASSET_DIR}/music.mp3", etc.
from pavo import render_video

render_video(
    "template.json",
    "output/final.mp4",
    variables={"ASSET_DIR": "/projects/my_video/assets"},
)

Missing keys raise immediately:

ValueError: Placeholder '${ASSET_DIR}' has no matching key in the provided variables

---

📱 Kick off Copilot coding agent tasks wherever you are with GitHub Mobile, available on iOS and Android.