Lov Media Acquisition

影视下载 · Media Download

Treat download as an observed job with recovery, not a fire-and-forget client action. Use aria2 as the primary transfer engine for direct and BitTorrent inputs. Treat qBittorrent as an optional adapter when its search, queue, inspection, or seeding capabilities add value.

Triggers

Activate when

• 用户说“开始下载，太慢就自动换源”“同时测试几个，选最快的”。
• 用户给出一个或多个 Magnet，希望下载完成并保留最优任务。
• The user asks to race sources, monitor a download, or recover a stalled transfer.

Do not activate when

• 用户还在比较内容不同的剪辑版本，尚未作出用户偏好决定。
• 目的磁盘的容量预检没有通过。
• The user asks only to search or rank releases without downloading.

Workflow (MANDATORY)

Step 0: Load safeguards and connection

Read $KIT_DIR/references/acquisition-policy.md and resolved configuration. Verify aria2 and assign an isolated job directory plus unique listen/RPC ports. If optional qBittorrent is enabled, verify its login, keep the WebUI on loopback, and query existing hashes before adding tasks. Mark those hashes as pre-existing; never delete, relocate, or retag them.

Step 1: Capacity preflight

Run scripts/storage_preflight.py with the decision file, destination, probe count, temporary probe allowance, and free-space reserve. Stop before transfer when the JSON result says ok=false; surface required, available, and shortfall.

Step 2: Create an isolated probe job

Use a unique tag and $OUTPUT/.media-fetch-probes/$JOB_ID. Probe no more than the configured concurrency. Keep each candidate in its own child directory so exact losing payloads can be cleaned without broad path deletion.

Step 3: Measure sustained usefulness

Run scripts/aria2_acquire.py for the selected inputs, using distinct job identities and ports when probes overlap. Ignore the warm-up window when comparing. Score sustained speed, availability, peers, progress, and ETA. A candidate with a brief burst followed by zero is weaker than a stable source. Keep advertised seeders and observed health in separate fields. Use scripts/qbittorrent_acquire.py only when that optional backend is enabled for this run.

Step 4: Continue and recover

• Move the winner to the final destination and keep polling until completion.

• Provide a user progress update at least once per minute while tools are active.

• If progress and traffic remain below thresholds for stall_seconds, pause the current task and resume the next proven candidate.

• Start or resume the default aria2 transfer with:

  python3 "$KIT_DIR/scripts/aria2_acquire.py" \
    --input INPUT --job-id JOB_ID --output-dir OUTPUT_DIR \
    --result ACQUISITION_JSON --watch --no-proxy
  

  Keep the .aria2 state in the isolated job directory and verify the final payload before relocation. For a direct URL with an opaque path, pass --output-name so the completed media retains a verifiable suffix.

• Switch to qBittorrent only after an enabled qBittorrent probe shows better sustained health or the user needs its queue/seeding behavior. Record the evidence and avoid a second full payload.

• If the complete candidate list is exhausted, preserve the best paused job, return to discovery for the next wave, preflight the incremental probe budget, and continue.

• Use a finite configured search wave per process; the agent owns the outer retry loop and remains in conversation with the user.

Step 5: Finish exact cleanup

After one task reports complete and its files exist, remove only tasks created by this job and exact isolated losing directories. Keep logs and the JSON acquisition report. Never delete by wildcard, parent directory, candidate name alone, or unresolved path.

Dependencies

aria2 1.36+ is required. Optional qBittorrent connection values come from QBITTORRENT_URL, QBITTORRENT_USERNAME, QBITTORRENT_PASSWORD, or the shared profile; secrets stay outside committed source.

Runtime context (shared)

运行前读取本 Skill 包的 skill.yaml，由宿主提供 skill-runtime/v1 上下文。字段解析顺序为：当前请求、项目上下文、个人 Preferences、品牌 Profile、通用默认值。

• 只使用 Manifest 声明的字段；Profile 保存公开品牌事实，Preferences 保存个人工作偏好。
• required: true 字段缺失时，按 Manifest 的问题配置向用户提出一个聚焦问题；用户明确同意后再保存回答。
• 报错提供可复制的 context_id、字段路径与来源，诊断内容避开秘密、完整私人路径和原始配置。