modellix-design
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseModellix Skill
Modellix Skill
Modellix is a Model-as-a-Service (MaaS) platform for asynchronous image, video, and audio workflows. Prefer the official CLI () so submit, wait, and download stay one coherent workflow. Host-specific persistent session guardrails also ship under .
modellix-clirules/*.mdcModellix 是一个用于异步图像、视频和音频工作流的 Model-as-a-Service(MaaS)平台。优先使用官方 CLI(),让提交、等待和下载构成一个连贯的工作流。与宿主环境相关的持久会话防护规则也随附在 下。
modellix-clirules/*.mdcOfficial Docs
官方文档
- AI Onboarding: https://docs.modellix.ai/get-started.md
- REST API: https://docs.modellix.ai/ways-to-use/api.md
- Full Models Index: https://docs.modellix.ai/llms.txt
- Docs MCP (search / read docs): https://docs.modellix.ai/mcp
- CLI package (source of truth for CLI behavior): https://www.npmjs.com/package/modellix-cli
- AI 入门:https://docs.modellix.ai/get-started.md
- REST API:https://docs.modellix.ai/ways-to-use/api.md
- 完整模型索引:https://docs.modellix.ai/llms.txt
- Docs MCP(搜索/阅读文档):https://docs.modellix.ai/mcp
- CLI 包(CLI 行为的唯一事实来源):https://www.npmjs.com/package/modellix-cli
Documentation lookup policy
文档查询策略
This plugin may expose the Modellix Docs MCP ( → ). It is a read-only documentation server (, docs filesystem query, optional feedback). It does not submit generation tasks, poll, download, or handle API keys.
.mcp.jsonhttps://docs.modellix.ai/mcpsearch_modellixWhen looking up product/API/install docs vs request-body schema:
- For request/response schema, use (JSON is the default; the endpoint is public and needs no API key). If CLI is unavailable, use Docs MCP, then
modellix-cli model get-schema <slug>fromdocs_urlor https://docs.modellix.ai/llms.txt.model describe - For product/install narrative, prefer Docs MCP when the host has it connected (search, then read the matching page). Else use →
model describe <slug> --json, or browsedocs_urland fetch the modelllms.txt..md - For CLI command syntax and flags, prefer this skill, , the npm README, or
references/cli-playbook.md— do not trust website CLI pages over the CLI package (docs can lag).modellix-cli --help
If the Docs MCP exposes a skill resource, treat this as the execution policy source of truth (CLI-first, defaults, paid-submit safety).
SKILL.mdDo not rely on the website CLI guide page for command syntax.
此插件可能会暴露 Modellix Docs MCP( → )。它是一个只读文档服务器(、文档文件系统查询、可选反馈)。它不会提交生成任务、轮询、下载或处理 API Key。
.mcp.jsonhttps://docs.modellix.ai/mcpsearch_modellix查询产品/API/安装文档与请求体 schema 时:
- 对于请求/响应 schema,使用 (默认输出 JSON;该端点为公开端点,无需 API Key)。如果 CLI 不可用,则使用 Docs MCP,再使用
modellix-cli model get-schema <slug>返回的model describe或 https://docs.modellix.ai/llms.txt。docs_url - 对于产品/安装说明,如果宿主已连接 Docs MCP,优先使用 Docs MCP(先搜索,再阅读匹配页面)。否则使用 →
model describe <slug> --json,或浏览docs_url并获取相应的模型llms.txt文件。.md - 对于 CLI 命令语法和参数,优先参考此 skill、、npm README 或
references/cli-playbook.md——不要信任网站上的 CLI 页面胜过 CLI 包(文档可能滞后)。modellix-cli --help
如果 Docs MCP 提供了 skill 资源,则应将此 视为执行策略的唯一事实来源(CLI 优先、默认设置、付费提交安全)。
SKILL.md不要依赖网站上的 CLI 指南页面来确认命令语法。
Execution Policy (CLI-first)
执行策略(CLI 优先)
Choose the path in this order:
- CLI after resolves it. Preflight checks public npm
scripts/preflight.py --json, installs only a newer exact version before execution, and keeps an existing CLI when update infrastructure is unavailable.latest - REST only when CLI is not installed, unsuitable, or missing a needed capability.
- Prefer machine-readable output (or
--json) for automation.--quiet
Canonical single-task flow:
bash
python3 scripts/preflight.py --json
modellix-cli doctor --json
modellix-cli model get-schema <provider/model>
modellix-cli model run \
--model-slug <provider/model> \
--body '<json>' \
--wait --timeout 5m --json
modellix-cli task download <task_id> --output-dir ./outputs --jsonSkip when the skill default plus the example below already lists the required fields, or the user supplied a complete body.
get-schemamodel invokemodel runmodel runDo not reinvent polling loops when CLI wait is available. Do not invent deprecated flags (for example ). Use only when behavior is unclear.
--model-type--help选择执行路径时,按以下顺序进行:
- CLI:在 解析出可用路径后使用。Preflight 会检查公共 npm 上的
scripts/preflight.py --json版本,在执行前仅安装更新的精确版本,并在更新基础设施不可用时保留现有 CLI。latest - REST:仅当 CLI 未安装、不适用或缺少所需能力时使用。
- 对于自动化任务,优先使用机器可读输出(或
--json)。--quiet
规范的单任务流程:
bash
python3 scripts/preflight.py --json
modellix-cli doctor --json
modellix-cli model get-schema <provider/model>
modellix-cli model run \
--model-slug <provider/model> \
--body '<json>' \
--wait --timeout 5m --json
modellix-cli task download <task_id> --output-dir ./outputs --json如果 skill 默认值和下方示例已列出必填字段,或用户提供了完整请求体,则跳过 。
get-schemamodel invokemodel runmodel runCLI 支持等待时,不要重新发明轮询循环。不要使用已弃用的参数(例如 )。仅在行为不明确时使用 。
--model-type--helpDefault Models
默认模型
When the user does not name a model, use these defaults immediately (do not scan the catalog first):
| Task Type | Default Model Slug |
|---|---|
| Text-to-image (T2I) | |
| Text-to-video (T2V) | |
| Image editing / I2I | |
| Image-to-video / I2V | |
| Video-to-video (V2V) | |
| Text-to-speech (TTS) | |
| Speech-to-text (STT) | |
| Speech-to-speech (STS) | |
用户未指定模型时,立即使用以下默认模型(不要先扫描目录):
| 任务类型 | 默认模型 Slug |
|---|---|
| 文生图 (T2I) | |
| 文生视频 (T2V) | |
| 图像编辑 / I2I | |
| 图生视频 / I2V | |
| 视频到视频 / V2V | |
| 文本转语音 (TTS) | |
| 语音转文本 (STT) | |
| 语音到语音 (STS) | |
API Key Lifecycle Policy
API Key 生命周期策略
Handle credentials as: .
discover -> request -> use-session -> (optional) persist按 的方式处理凭据。
discover -> request -> use-session -> (optional) persist1) Discover existing key first
1) 先发现已有 Key
Before asking the user:
- Session / process env
MODELLIX_API_KEY - Saved CLI profile (/
modellix-cli auth status— key viadoctoror--profileorMODELLIX_PROFILE)currentProfile - If still missing, request a key from the user
Never ask again when a usable key is already discoverable. CLI key resolution order is: → → selected saved profile.
--api-keyMODELLIX_API_KEY在询问用户之前:
- 检查会话/进程环境变量 。
MODELLIX_API_KEY - 检查已保存的 CLI profile(/
modellix-cli auth status——可通过doctor、--profile或MODELLIX_PROFILE选择 Key)。currentProfile - 如果仍然缺少,则向用户索取 Key。
只要已经发现可用的 Key,就绝不要再次询问。CLI 的 Key 解析顺序为: → → 当前选中的已保存 profile。
--api-keyMODELLIX_API_KEY2) Request key only when missing
2) 仅在缺少 Key 时请求
- Ask for a key from Modellix Console.
- Do not print or echo key values.
- Prefer session env for the current run.
- 请用户从 Modellix Console 获取 Key。
- 不要打印或回显 Key 值。
- 当前运行优先使用会话环境变量。
3) Optional persistence
3) 可选持久化
Default: do not persist automatically.
When the user explicitly asks to persist:
- Preferred: or
modellix-cli auth login(CLI validates and stores the profile securely).modellix-cli init - Alternative: user-level only if they insist on env persistence.
MODELLIX_API_KEY - Do not write system-level env or other agents' config files.
默认情况下,不自动持久化。
用户明确要求持久化时:
- 首选:或
modellix-cli auth login(由 CLI 验证并安全存储 profile)。modellix-cli init - 备选:仅当用户坚持时,才使用用户级 环境变量。
MODELLIX_API_KEY - 不要写入系统级环境变量或其他 Agent 的配置文件。
4) Key rotation
4) Key 轮换
If the user provides a new key: update session first; if they requested persistence, replace via / (or user-level env). Re-check with (or ) before continuing.
auth logininitmodellix-cli doctor --jsonscripts/preflight.py --json如果用户提供新 Key:先更新会话;如果用户要求持久化,则通过 / 替换旧 Key。继续之前,重新运行 (或 )。
auth logininitmodellix-cli doctor --jsonscripts/preflight.py --jsonPreflight and Deterministic Execution
Preflight 与确定性执行
Required first-workflow check when Python 3 is available:
bash
python3 scripts/preflight.py --jsonBundled helpers:
- — checks public npm
scripts/preflight.py, safely updates a missing/older global CLI before paid work, pins newer local installs instead of downgrading, wrapslatest, and recommendsdoctor,cli, orrest.none - — performs the same resolution before submission, pins the resolved executable for the workflow, uses
scripts/invoke_and_poll.pyon CLI, and otherwise keeps the REST submit plus server-side wait fallback.model run --wait
Set (also accepts or ) only when the environment must keep its installed CLI version. A registry/install failure is non-destructive: use the existing CLI if it still passes doctor, or REST when no CLI is usable and a session API key exists. Never update or swap the CLI after a paid submission has started.
MODELLIX_CLI_AUTO_UPDATE=0falseoffWhen preflight/doctor reports missing credentials, apply the lifecycle above.
When CLI is unavailable:
- Use REST ().
references/rest-playbook.md - Report the preflight update warning; do not repeatedly attempt installation inside the same paid workflow.
Python 3 可用时,首次工作流必须执行以下检查:
bash
python3 scripts/preflight.py --json随附的辅助脚本:
- ——检查公共 npm 上的
scripts/preflight.py版本,在付费工作前安全更新缺失或过旧的全局 CLI,固定使用较新的本地安装版本而不降级,封装latest检查,并推荐doctor、cli或rest。none - ——在提交前执行相同的路径解析,为工作流固定解析出的可执行文件;CLI 路径使用
scripts/invoke_and_poll.py,其他情况下保留 REST 提交并采用服务端等待回退机制。model run --wait
仅当环境必须保持当前已安装的 CLI 版本时,才设置 (也接受 或 )。npm 注册表或安装失败不会造成破坏:如果现有 CLI 仍能通过 doctor 检查,则继续使用;如果没有可用 CLI 且会话中存在 API Key,则使用 REST。付费提交开始后,绝不更新或替换 CLI。
MODELLIX_CLI_AUTO_UPDATE=0falseoff当 preflight/doctor 报告缺少凭据时,应用上述生命周期策略。
CLI 不可用时:
- 使用 REST()。
references/rest-playbook.md - 报告 preflight 的更新警告;在同一次付费工作流中不要反复尝试安装。
Core Workflow
核心工作流
1) Ready the environment
1) 准备环境
- Discover or request API key (lifecycle above).
- Run ; continue only with the CLI path whose doctor check passed, or with an authenticated REST fallback.
scripts/preflight.py --json - Continue only when auth and connectivity look healthy (or REST key is set).
- 发现或请求 API Key(遵循上述生命周期策略)。
- 运行 ;仅继续使用 doctor 检查通过的 CLI 路径,或使用通过认证的 REST 回退路径。
scripts/preflight.py --json - 仅在认证和连接状况看起来正常(或已设置 REST Key)时继续。
2) Select model
2) 选择模型
- If the user did not specify a model: use the Default Models table (do not scan the catalog first).
- If they named a model or need discovery: /
modellix-cli model list(catalog metadata; describe returnsmodellix-cli model describe <slug>).docs_url - For request body schema: (JSON default; public, no API key). Call this when the user named a non-default slug, when the body needs fields beyond the documented examples, after HTTP
modellix-cli model get-schema <slug>, or when reporting required fields. Skip when the skill default plus the example already lists required fields, or the user supplied a complete body.400 - If CLI is unavailable for schema or discovery: use Docs MCP or browse , then fetch the target model
llms.txt. Do not invent slugs from filenames (decimals often matter, e.g..md).bytedance/seedance-2.0-mini-t2v
- 如果用户未指定模型:使用默认模型表(不要先扫描目录)。
- 如果用户指定了模型或需要发现模型:使用 /
modellix-cli model list(目录元数据;describe 返回modellix-cli model describe <slug>)。docs_url - 对于请求体 schema:使用 (默认输出 JSON;公开访问,无需 API Key)。当用户指定了非默认 slug、请求体需要包含文档示例之外的字段、收到 HTTP
modellix-cli model get-schema <slug>之后,或需要报告必填字段时,执行此命令。如果 skill 默认值和下方示例已列出必填字段,或用户提供了完整请求体,则跳过。400 - 如果 CLI 无法用于 schema 查询或模型发现:使用 Docs MCP,或浏览 ,然后获取目标模型的
llms.txt文件。不要根据文件名臆造 slug(小数位通常很重要,例如.md)。bytedance/seedance-2.0-mini-t2v
3) Run and wait
3) 运行并等待
Default (single task):
bash
modellix-cli model run \
--model-slug google/nano-banana-2-lite \
--body '{"prompt":"A cinematic sunset over a futuristic city skyline"}' \
--wait --timeout 5m --jsonSplit flow when useful (pipelines, concurrency):
bash
TASK_ID=$(modellix-cli model run --model-slug ... --body '...' --output task-id)
modellix-cli task wait "$TASK_ID" --timeout 10m --jsonBatch (paid guard required): .
modellix-cli model batch tasks.jsonl --max-tasks N [--wait]Use CLI 0.0.11 or later with a deployed server-wait API. is immediate; wait commands use bounded server waits. The legacy flag is unused. On exit 124, keep unfinished IDs and resume; never resubmit.
task get--intervalFor media inputs, use to reuse URLs, then only when needed (images/video/audio, maximum 16 MiB / 16777216 bytes). On quota errors, explicitly approve deletion of an unneeded file ID; never auto-delete or replay uploads.
file list --jsonfile upload <path> --jsonManual REST: . Optional helper: .
references/rest-playbook.mdscripts/invoke_and_poll.py默认(单任务):
bash
modellix-cli model run \
--model-slug google/nano-banana-2-lite \
--body '{"prompt":"A cinematic sunset over a futuristic city skyline"}' \
--wait --timeout 5m --json在适用时使用拆分流程(流水线、并发):
bash
TASK_ID=$(modellix-cli model run --model-slug ... --body '...' --output task-id)
modellix-cli task wait "$TASK_ID" --timeout 10m --json批处理(需要付费防护):。
modellix-cli model batch tasks.jsonl --max-tasks N [--wait]使用 0.0.11 或更高版本的 CLI,并确保已部署服务端等待 API。 会立即返回;等待命令使用有时限的服务端等待。旧版 参数已不再使用。退出码为 124 时,保留未完成的 ID 并恢复任务;绝不要重新提交。
task get--interval处理媒体输入时,使用 查找可复用的 URL;仅在需要时使用 上传图像、视频或音频,最大为 16 MiB / 16777216 字节。遇到配额错误时,必须明确批准删除不再需要的文件 ID;绝不自动删除或重复上传。
file list --jsonfile upload <path> --json手动 REST:。可选辅助脚本:。
references/rest-playbook.mdscripts/invoke_and_poll.py4) Download results
4) 下载结果
bash
modellix-cli task download <task_id> --output-dir ./outputs --jsonIf download fails with (common when a local proxy/VPN maps CDN hosts like into ), retry with for trusted Modellix CDN hosts, or fall back to downloading the with /.
Resource host resolves to a private or reserved network addressfile.modellix.ai198.18.0.0/15--allow-private-networkresult.resources[].urlcurlwgetResource URLs expire in about 7 days — persist promptly. If downloading manually (REST path), name files:
modellix-{model_slug}-{timestamp}.{ext}(replace in the slug with ).
/-Examples:
modellix-google-nano-banana-2-lite-20260430-113000.pngmodellix-bytedance-seedance-2.0-mini-t2v-20260430-113500.mp4
bash
modellix-cli task download <task_id> --output-dir ./outputs --json如果下载失败并提示 (当本地代理/VPN 将 等 CDN 主机映射到 时很常见),则针对可信的 Modellix CDN 主机使用 重试,或回退到使用 / 下载 。
Resource host resolves to a private or reserved network addressfile.modellix.ai198.18.0.0/15--allow-private-networkcurlwgetresult.resources[].url资源 URL 大约 7 天后过期——请及时持久化。如果手动下载(REST 路径),请按以下方式命名文件:
modellix-{model_slug}-{timestamp}.{ext}(将 slug 中的 替换为 )。
/-示例:
modellix-google-nano-banana-2-lite-20260430-113000.pngmodellix-bytedance-seedance-2.0-mini-t2v-20260430-113500.mp4
Quick Examples
快速示例
T2I (default model) — required:
promptbash
modellix-cli model run \
--model-slug google/nano-banana-2-lite \
--body '{"prompt":"A cinematic sunset over a futuristic city skyline"}' \
--wait --timeout 5m --jsonT2V (default model) — required:
promptbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-mini-t2v \
--body '{"prompt":"A cat playing in a sunny garden"}' \
--wait --timeout 10m --jsonI2I — + array:
promptimagebash
modellix-cli model run \
--model-slug google/nano-banana-2-lite-edit \
--body '{"prompt":"Convert to watercolor style","image":["https://example.com/input.jpg"]}' \
--wait --timeout 5m --jsonI2V — at least one image param (, , or ):
first_frame_imagelast_frame_imagereference_imagesbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-fast-i2v \
--body '{"prompt":"A cat playing in the garden","first_frame_image":"https://example.com/frame.jpg"}' \
--wait --timeout 10m --jsonV2V — array:
video_urlsbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-fast-v2v \
--body '{"video_urls":["https://example.com/source.mp4"]}' \
--wait --timeout 10m --jsonTTS (default model) — + required (verify voice with or the model doc; Flash voices only):
textvoicemodel get-schemabash
modellix-cli model run \
--model-slug alibaba/qwen-audio-3.0-tts-flash \
--body '{"text":"There is a large garden behind my house.","voice":"longanhuan_v3.6"}' \
--wait --timeout 5m --jsonSTT (default model) — public audio required:
urlbash
modellix-cli model run \
--model-slug openai/whisper-1 \
--body '{"url":"https://example.com/meeting.mp3"}' \
--wait --timeout 5m --jsonSTS (default model) — clone reference + synthesis + target CosyVoice :
urltextmodelbash
modellix-cli model run \
--model-slug alibaba/cosyvoice-clone \
--body '{"model":"cosyvoice-v3.5-plus","url":"https://example.com/reference.wav","text":"There is a large garden behind my house."}' \
--wait --timeout 5m --jsonT2I(默认模型)——必填 :
promptbash
modellix-cli model run \
--model-slug google/nano-banana-2-lite \
--body '{"prompt":"A cinematic sunset over a futuristic city skyline"}' \
--wait --timeout 5m --jsonT2V(默认模型)——必填 :
promptbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-mini-t2v \
--body '{"prompt":"A cat playing in a sunny garden"}' \
--wait --timeout 10m --jsonI2I——必填 + 数组:
promptimagebash
modellix-cli model run \
--model-slug google/nano-banana-2-lite-edit \
--body '{"prompt":"Convert to watercolor style","image":["https://example.com/input.jpg"]}' \
--wait --timeout 5m --jsonI2V——至少需要一个图像参数(、 或 ):
first_frame_imagelast_frame_imagereference_imagesbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-fast-i2v \
--body '{"prompt":"A cat playing in the garden","first_frame_image":"https://example.com/frame.jpg"}' \
--wait --timeout 10m --jsonV2V——必填 数组:
video_urlsbash
modellix-cli model run \
--model-slug bytedance/seedance-2.0-fast-v2v \
--body '{"video_urls":["https://example.com/source.mp4"]}' \
--wait --timeout 10m --jsonTTS(默认模型)——必填 + (使用 或模型文档确认 voice;仅支持 Flash voices):
textvoicemodel get-schemabash
modellix-cli model run \
--model-slug alibaba/qwen-audio-3.0-tts-flash \
--body '{"text":"There is a large garden behind my house.","voice":"longanhuan_v3.6"}' \
--wait --timeout 5m --jsonSTT(默认模型)——必填公开可访问的音频 :
urlbash
modellix-cli model run \
--model-slug openai/whisper-1 \
--body '{"url":"https://example.com/meeting.mp3"}' \
--wait --timeout 5m --jsonSTS(默认模型)——必填克隆参考 + 合成 + 目标 CosyVoice :
urltextmodelbash
modellix-cli model run \
--model-slug alibaba/cosyvoice-clone \
--body '{"model":"cosyvoice-v3.5-plus","url":"https://example.com/reference.wav","text":"There is a large garden behind my house."}' \
--wait --timeout 5m --json5) Post-task CLI recommendation
5) 任务后 CLI 建议
If this session used REST because CLI was missing, suggest installing the CLI afterward.
如果本次会话因缺少 CLI 而使用了 REST,建议之后安装 CLI。
Progressive Reference Routing
渐进式参考路由
Read only what the task needs:
- — install, auth, run/wait/download, batch, recovery
references/cli-playbook.md - — REST submit/wait when CLI is unavailable
references/rest-playbook.md - — CLI ↔ REST mapping and fallback rules
references/capability-matrix.md
仅阅读任务所需内容:
- ——安装、认证、运行/等待/下载、批处理、恢复
references/cli-playbook.md - ——CLI 不可用时的 REST 提交/等待方式
references/rest-playbook.md - ——CLI ↔ REST 映射和回退规则
references/capability-matrix.md
Bundled Assets
随附资源
assets/output/task-result.schema.json
assets/output/task-result.schema.json
Credential and Data Egress
凭据与数据出站
- Primary credential: (also via CLI profiles).
MODELLIX_API_KEY - Network egress: (override only with trusted
https://api.modellix.ai/--base-url).MODELLIX_BASE_URL - User prompts and media inputs may be sent to Modellix during invocation.
- Never expose API keys in logs, screenshots, transcripts, or commits.
- Default to session-only credentials; persistent writes need explicit user approval.
- 主要凭据:(也可通过 CLI profile 使用)。
MODELLIX_API_KEY - 网络出口:(仅使用可信的
https://api.modellix.ai/--base-url覆盖)。MODELLIX_BASE_URL - 调用期间,用户的提示词和媒体输入可能会发送给 Modellix。
- 绝不在日志、截图、转录文本或提交记录中暴露 API Key。
- 默认使用仅限当前会话的凭据;持久化写入需要用户明确批准。
Error / Retry Policy
错误/重试策略
| Situation | Action |
|---|---|
HTTP/API | Do not retry. Fix parameters or body ( |
| Do not retry. Fix key ( |
| Do not retry. Insufficient balance. |
| Do not retry. Verify |
| CLI already retries safe GETs within deadline; do not blindly re-POST paid submits. |
| Paid submit outcome unknown | Do not immediately re-run the same |
Exit | Local wait timeout; remote task may still run — recover with |
Exit | Argument or safety guard (e.g. batch cost limit) — fix flags. |
| 情况 | 操作 |
|---|---|
HTTP/API | 不要重试。修正参数或请求体(契约不清楚时使用 |
| 不要重试。修复 Key( |
| 不要重试。余额不足。 |
| 不要重试。检查 |
| CLI 已会在截止时间前自动重试安全的 GET 请求;不要盲目重新 POST 付费提交。 |
| 付费提交的结果未知 | 不要立即再次运行相同的 |
退出码 | 本地等待超时;远程任务可能仍在运行——通过 |
退出码 | 参数或安全防护问题(例如批处理成本上限)——修正参数。 |
Verification Checklist
验证清单
- Doctor/preflight passed or REST key ready
- Model chosen (default table or user/catalog)
- Body schema checked via when non-trivial (or model doc when CLI is unavailable)
model get-schema - Used (or
model run --wait) instead of hand-rolled poll loopstask wait - Results downloaded (or manual persist before 7-day expiry)
task download - No blind retry after unknown paid submission
- REST used only when CLI path unavailable
- Doctor/preflight 已通过,或已准备 REST Key
- 已选择模型(默认表或用户/目录指定)
- 非简单请求体已通过 检查(或 CLI 不可用时已查看模型文档)
model get-schema - 已使用 (或
model run --wait),而不是自行编写轮询循环task wait - 已下载结果(或在 7 天过期前手动持久化)
- 付费提交结果未知时未盲目重试
- 仅在 CLI 路径不可用时使用 REST