video-analysis: mode parameter (default both), sound-design brief on the result; sonilo 0.19.0, sonilo-cli 0.18.0 - #44
Merged
Conversation
…the result; release sonilo 0.19.0 and sonilo-cli 0.18.0
POST /v1/video-analysis takes an optional `mode`: both (the server default),
music or sfx. Both submit() and analyze() on the sync and async resources
accept it and pass it straight through; it is omitted when unset so the
server default applies, and its values are not checked client-side, like
variants_num.
VideoAnalysisResult gains `mode` (an echo of the request), `sfx_segments`
(shot-sized sections, label always "none") and `sfx_prompt` (one string, the
sound design is authored once regardless of variants_num). The parser
sanitizes sfx_segments the way it already sanitizes segments, and reads a
blank or non-string sfx_prompt as None. mode=music reproduces the previous
result shape plus the echo.
The CLI adds `--mode {both,music,sfx}` and emits mode, sfx_segments and
sfx_prompt in its JSON brief when the task carries them; a music or sfx
brief keeps the shape it always had.
The video-analysis duration cap is now 480 seconds (was 360); the price is
unchanged. README, the CLI README and context7.json say so.
sonilo 0.18.2 -> 0.19.0, sonilo-cli 0.17.2 -> 0.18.0, and the CLI's core pin
widens to sonilo>=0.19.0,<0.20. The CLI's uv.lock still resolves the
published 0.18.1 and needs a refresh once 0.19.0 is on PyPI.
Lightsage docs evalsWaiting for the staging docs URL before running evals. Lightsage will start the selected PR evals automatically when GitHub reports a successful docs deployment for this PR. This usually happens within 15 minutes. Commit: |
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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Mirrors the video-analysis change now live on the API.
SDK (
sonilo0.18.2 → 0.19.0)client.video_analysis.submit()/analyze()(sync and async) acceptmode—"both"(the server default),"music"or"sfx". Sent only when set; values are not checked client-side, same rule asvariants_num.VideoAnalysisResultgainsmode,sfx_segments(same shape assegments,labelalways"none") andsfx_prompt(one string — authored once regardless ofvariants_num). The parser coercessfx_segmentswith the same sanitizer assegments; a blanksfx_promptreads asNone.mode="music"reproduces the previous result exactly.context7.json.CLI (
sonilo-cli0.17.2 → 0.18.0)sonilo video-analysis --mode both|music|sfx. The JSON brief now includesmode, andsfx_segments/sfx_promptwhen present, in the documented key order.sonilo>=0.19.0,<0.20.sonilo-cli/uv.lockis deliberately not refreshed: it cannot resolve untilsonilo0.19.0 is on PyPI, and CI does not read the lockfiles.Verified
pytest -q→ 332 passedpytest -q→ 172 passedpip checkclean after the pin change.Release order
Publish
sonilo0.19.0 first, thensonilo-cli0.18.0.