Runninghub

Billing / Reporting

RunningHub scripts may expose two billing schemes. Treat them as complementary, not mutually exclusive:

A. Preflight billing check (PREFLIGHT_BILLING_MODE)

Use this when the script prints a pre-run estimate / charging mode before execution.

• Read PREFLIGHT_BILLING_MODE if present.
• Report it before you say the job has started or before you present final results.
• Explain it in human language:
  • unknown: 目前还看不出这次会怎么计费，先帮你跑，跑完再回报实际扣费情况。
  • balance: 这次预计会从余额扣款。
  • coins: 这次预计会扣点数 / coins。
  • mixed: 这次可能同时涉及余额与点数 / coins。
• If preflight fields are absent, do not invent them.

B. Post-run billing report (BILLING_MODE, BALANCE_*, COINS_*)

After every completed run, you MUST inspect and report billing outcome from script output.

Required reads after task completion:

• BILLING_MODE
• BALANCE_DELTA
• COINS_DELTA

Also read related fields when present (for clearer wording):

• BALANCE_BEFORE, BALANCE_AFTER
• COINS_BEFORE, COINS_AFTER

Human-language interpretation:

• unknown: 这次执行成功了，但脚本没有给出明确计费模式；若有 delta 就一起说明实际变化。
• balance: 明确说是扣余额；优先说明 BALANCE_DELTA，有 before/after 就一起说。
• coins: 明确说是扣点数 / coins；优先说明 COINS_DELTA，有 before/after 就一起说。
• mixed: 明确说同时涉及余额与点数 / coins；把两边 delta 都讲清楚。

Assistant Reporting Format

Billing 回报顺序是硬性规则：

1. If preflight exists, say it first — 先告知这次预计怎么计费，再继续执行说明。
2. After completion, always report final billing — 不管用户有没有追问，都要读取并回报 BILLING_MODE / BALANCE_DELTA / COINS_DELTA。
3. Use human wording, not raw field dump — 可以引用数值，但不要只贴变量名。
4. If COST:¥X.XX is present, report it naturally too — 例如「这次花了 ¥0.50，另外余额少了 0.50」；若两者语义重复，也要优先确保 billing mode + delta 有讲清楚。

Suggested phrasing:

• Preflight: 先跟你说一下，这次预计会从余额扣款，我这就开始跑～
• Post-run / balance: 搞定啦～ 这次是扣余额，余额减少 0.50；如果脚本有 before/after，就一起带上。
• Post-run / coins: 搞定啦～ 这次是扣点数，coins 少了 20。
• Post-run / mixed: 搞定啦～ 这次同时动用了余额和点数：余额少了 0.50，coins 少了 20。
• Post-run / unknown: 搞定啦～ 这次脚本没明确标出计费模式，不过从结果看余额少了 0.50 / coins 少了 20。

API Key Setup

When user needs to set up or check their API key → Read {baseDir}/references/api-key-setup.md and follow its instructions.

Quick check: python3 {baseDir}/scripts/runninghub.py --check

Routing Table

AI Application

When user mentions "AI应用", "workflow", "webappId", pastes a RunningHub AI app link, or asks to browse/discover apps ("有什么应用", "最热门的", "最新的", "推荐什么") → Read {baseDir}/references/ai-application.md and follow its complete flow.

Script Usage

Execution flow for ALL generation tasks:

1. Slow tasks (video / 3D / music / AI app): First send message notification → "开始生成啦，一般需要 X 分钟，请稍等～" → then exec the script
2. Fast tasks (image / TTS / upscale): Directly exec the script (notification optional)

python3 {baseDir}/scripts/runninghub.py \
  --endpoint ENDPOINT \
  --prompt "prompt text" \
  --param key=value \
  -o /tmp/openclaw/rh-output/name_$(date +%s).ext

Optional flags: --image PATH, --video PATH, --audio PATH, --param key=value (repeatable) Discovery: --list [--type T], --info ENDPOINT

Example — text to image:

python3 {baseDir}/scripts/runninghub.py \
  --endpoint rhart-image-n-pro/text-to-image \
  --prompt "a cute puppy, 4K cinematic" \
  --param resolution=2k --param aspectRatio=16:9 \
  -o /tmp/openclaw/rh-output/puppy_$(date +%s).png

Output

For media delivery and error handling details → Read {baseDir}/references/output-delivery.md.

Key rules (always apply):

• ALWAYS call message tool to deliver media files, then respond NO_REPLY.
• For video outputs, scripts now validate downloaded files with ffprobe (must read duration + stream info). If the first download is corrupt, scripts delete the bad file and retry up to 2 times automatically.
• If video download validation still fails after retries, treat it as a failure (not success). Read the emitted error JSON and surface the raw download_url / source URL to the user or upper layer as fallback information.
• If message fails, retry once. If still fails, include OUTPUT_FILE:<path> and explain.
• Print text results directly. Include cost if COST: line present.