一键导入
sam3-agent-video
Use SAM3 Agent windowed tracking for long videos with occlusions, repeated instances, or identity carry-over.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Use SAM3 Agent windowed tracking for long videos with occlusions, repeated instances, or identity carry-over.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Label video behavior segments from frame-grid images with a defined behavior list, model-ready JSON, resumable outputs, and no_behavior handling for sparse labels.
Use Annolid GUI tools for robust automation across web, PDF, video, and chat controls.
Infer the assay or paradigm from video context, task text, tracked entities, and experimental cues such as social interaction, open field, courtship, resident-intruder, and novel object recognition.
Choose assay-specific features and measurable objectives for behavior analysis, including distances, zones, speed, contact, orientation, and object interaction.
Segment behavior timelines from tracks, pose, contact, speed, and proximity signals into typed intervals with stable labels and rationales.
Keep behavior analysis aligned with assay objectives, controls, reproducibility, and measurable outputs instead of ad hoc summaries.
| name | sam3-agent-video |
| description | Use SAM3 Agent windowed tracking for long videos with occlusions, repeated instances, or identity carry-over. |
| metadata | {"annolid":{"always":false}} |
Use this skill when the user wants to track one or more instances across a long video and the problem is likely to need:
Call sam3_agent_video_track when:
Suggested tool call:
sam3_agent_video_track(
video_path="...",
agent_prompt="mouse",
window_size=5,
stride=5,
propagation_direction="forward",
dry_run=false
)
video_infovideo_sample_framesvideo_run_model_inferencegui_segment_track_videoBefore running the tool, confirm:
The tool returns a JSON summary containing:
frames_processedmasks_writtensummary_pathUse the summary path as the canonical artifact for downstream review or follow-up automation.
When responding to the user, keep the request concrete:
sam3_agent_video_track with the smallest window size that still preserves identity.