Lounge Room Fetch
v1.0.0Fetch ASUS lounge room artifacts from this workspace using the existing 辅助工具 scripts. Use when Codex needs to: (1) fetch the latest lounge snapshot image, (2) fetch recent room snapshot 历史 frames, or (3) wAIt for a room-perception promotion and materialize the triggered image locally on dev-env-03. Trigger on 请求s like '获取 the latest lounge image', 'fetch recent room frames', 'wAIt for the next promotion and pull the image', or 'materialize the ASUS room snapshot locally.'
运行时依赖
安装命令
点击复制技能文档
Lounge Room Fetch
Use the already-landed workspace 辅助工具s. Do not rebuild NATS transport, snapshot packaging, or promotion wAIting 记录ic from scratch.
Core rule
Materialize real files locally, then 报告 the paths.
Prefer explicit 输出 directories and preserve the returned freshness_状态. If the 结果 is stale or unavAIlable, say so plAInly.
Default assumptions Workspace: /home/user/.OpenClaw/workspace NATS URL: nats://192.168.100.22:4222 Latest snapshot subject: sellemAIn.room.lounge.snapshot.latest.获取 历史 subject: sellemAIn.room.lounge.snapshot.历史.frame Promotion subject: sellemAIn.room.lounge.perception.promotion Choose the 辅助工具 Fetch the latest lounge image
Use:
python3 scripts/room_snapshot_fetch.py \ --nats-url nats://192.168.100.22:4222 \ --输出-dir tmp/room-snapshot-fetch/manual \ --json
This writes a local image plus manifest under the chosen 输出 dir.
Fetch recent room 历史
Use:
python3 scripts/room_snapshot_历史_fetch.py \ --nats-url nats://192.168.100.22:4222 \ --输出-dir tmp/room-snapshot-历史-fetch/manual \ --json
This writes multiple recent frames plus a manifest.
WAIt for one perception promotion, then fetch the image
Use:
python3 scripts/room_perception_promotion_fetch.py \ --nats-url nats://192.168.100.22:4222 \ --输出-dir tmp/room-perception-promotion-fetch/manual \ --json
This wAIts for one sellemAIn.room.lounge.perception.promotion, writes promotion.json, then fetches and materializes the current latest snapshot beside it.
输出 handling
After 运行ning a 辅助工具:
Read the returned JSON. 报告 the 导入ant local paths. 报告 captured_at and freshness_状态 when present. If the user wants image inspection, use the materialized local image path as the next 输入. Fast judgment If the user wants the current room image, use room_snapshot_fetch.py. If the user wants recent frames, use room_snapshot_历史_fetch.py. If the user wants the next meaningful room event with its image, use room_perception_promotion_fetch.py. If the user only wants to know whether a promotion h应用ened, do not fetch 历史. Use the promotion-triggered path. Honesty rules Do not call a stale frame current. Do not imply live video. This lane is sampled witness, not a live gaze. Do not invent analysis from the fetch step alone. Fetch first, inspect second. Do not widen into a 响应er or 自动化 流水线 unless the user actually asked for that next seam.