| name | kuavi-deep-analyze |
| description | Deep multi-pass video analysis with parallel sharding and zoom |
| agent | video-analyst |
| context | fork |
| argument-hint | <video-path> <question> |
| disable-model-invocation | true |
KUAVi Deep Analyze
Perform comprehensive, multi-pass video analysis with parallel temporal sharding and progressive zoom.
Instructions
- Parse
$ARGUMENTS for: video path and question/task.
- Index: If the video is not yet indexed, call
kuavi_index_video.
- Orient: Call
kuavi_get_scene_list to understand the video structure.
- Parallel Shard Analysis: Call
kuavi_analyze_shards with the question to get per-shard answers from parallel LLM workers. Use this to identify which temporal regions are most relevant.
- Pass 1 — Visual Search: Use
kuavi_search_video across multiple fields (summary, action, visual, temporal) to find relevant segments. Cross-reference with shard analysis results.
- Pass 2 — Transcript Search: Use
kuavi_search_transcript and kuavi_get_transcript to find spoken evidence. Note any discrepancies with visual findings.
- Pass 3 — Cross-Reference: For key findings, use the 3-pass zoom protocol:
- Level 1 zoom to locate
- Level 2 zoom to read
- Level 3 zoom to confirm specific values
- Pixel Analysis: When counting, measuring, or comparing, use pixel tools (
crop_frame, diff_frames, threshold_frame, frame_info) for deterministic evidence.
- Code Reasoning: Use
kuavi_eval for complex computations, iterating over frames, or chaining multiple operations.
- Budget Check: Monitor
kuavi_get_session_stats — synthesize after 15 tool calls or 120 seconds.
- Synthesize: Combine all evidence into a comprehensive, timestamped answer.
Arguments
The first argument is the video path. Remaining text is the question.
Example: /kuavi-deep-analyze /path/to/video.mp4 What are the key events and when do they occur?