Gh Stack
remotion-dev/remotion
Manages stacked PRs and splits multi-part work into reviewable branches with gh-stack.
火山方舟 Ark 图片/视频生成入口:支持 profile 默认资源与临时 API Key/Base URL/Endpoint;显式 Endpoint 不受当前 plan profile 误导。图片同步返回,视频异步轮询。
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install volcengine/ark-cli arkcli-gen --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/arkcli-gen .claude/skills/arkcli-gen && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .claude/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-genType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install volcengine/ark-cli arkcli-gen --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/arkcli-gen .agents/skills/arkcli-gen && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .agents/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install volcengine/ark-cli arkcli-gen --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/arkcli-gen .cursor/skills/arkcli-gen && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .cursor/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/volcengine/ark-cli.git --path skills/arkcli-gen--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install volcengine/ark-cli arkcli-gen --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/arkcli-gen .gemini/skills/arkcli-gen && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .gemini/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install volcengine/ark-cli arkcli-genInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/arkcli-gen .github/skills/arkcli-gen && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .github/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add volcengine/ark-cli --skill arkcli-gen -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install volcengine/ark-cli arkcli-gen --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/volcengine/ark-cli.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/arkcli-gen .opencode/skills/arkcli-gen && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "arkcli-gen" agent skill from https://github.com/volcengine/ark-cli/tree/main/skills/arkcli-gen into .opencode/skills/arkcli-gen/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "arkcli-gen", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
arkcli-gen火山方舟 Ark 图片/视频生成入口:支持 profile 默认资源与临时 API Key/Base URL/Endpoint;显式 Endpoint 不受当前 plan profile 误导。图片同步返回,视频异步轮询。
Arkcli Gen is an agent skill from volcengine/ark-cli. 火山方舟 Ark 图片/视频生成入口:支持 profile 默认资源与临时 API Key/Base URL/Endpoint;显式 Endpoint 不受当前 plan profile 误导。图片同步返回,视频异步轮询。
Its SKILL.md is about 3.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 6 other files, including reference files (for example `references/arkcli-gen.md`, `references/evals.md` and `references/gen-meta.md`).
It sits in Media & Creative. The repository describes itself as: The fastest way to put Volcengine Ark in your terminal and your AI agent — go from prompt to generated media, multimodal answer, or deployed endpoint in a single command, no API… The licence is Apache-2.0.
5 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit fb5b7be. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Shell commands in SKILL.md call:
jqFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Arkcli Gen loads about 3.9k tokens when it runs, and up to ~18k if it reads all its reference files. Until then it costs about 31 tokens; SKILL.md has 1,068 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from volcengine/ark-cli at commit fb5b7be, republished under its Apache-2.0 licence (© volcengine). 1,068 words, ~3,907 tokens.
.claude/skills/arkcli-gen/SKILL.md (or your agent's skills folder). This skill also uses 5 other files; get the full folder from GitHub.CRITICAL — 开始前 MUST 先用 Read 工具读取 ../arkcli-shared/SKILL.md(认证闸门、模型查找回退、共享安全规则)。
CRITICAL — 真实生成是工作流,不是猜一条命令:按 Step 1 → Step 2 → 必要时 Step 2.5 → Step 3 执行。用户只要求 --dry-run 时例外:全程本地,不先跑在线 Resources / Models / Usage 准入;在线未知项保留 unresolved。执行前务必读 references/arkcli-gen.md。
CRITICAL — 用户显式给出 API Key / Base URL / Endpoint 时,MUST 先读 ../arkcli-shared/references/execution-context.md。显式 Endpoint 的权威元数据优先于当前 profile。
火山额外约束:不要因为 active profile 是 Agent/Coding Plan 就把用户给出的 Endpoint 当套餐模型调用。
先把用户要求拆成可验收项,再按资源、能力、整批预算、执行、成品检查推进。执行前读取
references/intent-and-validation.md:它约定不完整意图的默认值、
参数组合、凭证错误分流和成品验收。参数表是候选能力,不代表每个模型都支持。
用户已明确要图片或视频时,生成命令显式带 --modality image|video;不要让名称解析覆盖用户意图。
只使用用户本轮提供或明确授权复用的素材;不得从历史目录、旧任务或相似文件名擅自加入额外参考图/视频。
生成模型能出图,不代表驱动当前 Agent 的宿主模型能看图。Read 工具描述说支持图片, 也不是当前宿主模型的视觉能力证明。 当前会话没有可靠的视觉能力声明或已验证的兼容性时, 按“未准入”处理:不要原生 Read 图片、视频或抽出的帧,也不要用成品试探能否读取。 本 Skill、参考说明和 JSON 等文本仍可正常 Read。
arkcli-chat 或有固定产出形态的 arkcli-understand。用户说"生成一个视频/一张图",本质是至少三件独立的事,必须按序;批量或多阶段任务还要先确认整批可完成:
① 本次资源从哪里来 ── 用户显式 Endpoint 优先;否则看当前 profile
② 该模型支持哪些参数 ── 不查就传参 = 瞎猜 = 被校验拒/被后端拒
②.5 多候选/多阶段额度与任务数 ── 先算完整批次,额度已耗尽就不启动半批任务
③ 按可用参数真去生成 ── 每个请求只提交一次并立即保存 task_id把这三步压成"直接 +gen 猜一条命令",正是失败之源:模型名形态不对会 404,参数模型不支持会被拒。
+gen 的生产调用按以下固定优先级解析能力:
explicit --modality > output_modalities > task types > unknownoutput_modalities;缺失时再读取 FoundationModel 的 task_types / filter_task_types。ep-* 时,先读取 Endpoint 的 ModelReference.FoundationModel(name, version),再精确匹配同版本模型的上述结构化元数据。seedream、seedance 或任何国内/海外品牌前缀推断 image/video。unknown,提示用户显式传 --modality image|video;禁止静默猜测。+gen --dry-run 是纯本地 Client Preview:不读取 Endpoint/模型元数据、不调用
生成 API,也不下载或打开文件。显式 --modality 最可靠;已知
seedream/seedance 模型名可本地判断,其他模型或 Endpoint 必须显式传
--modality image|video。在线才能补齐的执行上下文会以 unresolved 和
fidelity=partial 明示。用户意图: "生成 X"
│
▼ Step 1【强制】解析本次资源
│ 用户给 Endpoint → arkcli resources resolve <ep-id>
│ 未给 Endpoint → arkcli resources list --modality image|video
│
│ 当前 profile 可用资源:
│ platform → 列 EP (ep-xxx) ┐
│ agent-plan → 列视觉模型名 ├─ 选一个,记为 $MODEL
│ coding-plan → 列 EP (借道 platform) ┘
│
▼ Step 2【强制】查可用参数 ──► models get(EP 用 resolve 得到的绑定模型查能力)
│ 模型名 + 有 sp → **只能**用列出的参数,取值落 min/max/enum 内
│ 模型名 + sp 空(未配置或当前不可解析) → +gen 自动套 modality 兜底默认(video 720p/5s; image 不填 size)
│ EP(ep-xxx) → 可解析绑定则查精确版本;不可解析则说明未知,不猜支持
│
▼ Step 2.5【批量/多阶段】额度预检 ──► plan/free-quota 快照;记录完整 create 数
│
▼ Step 3 据可用参数生成 ──► arkcli +gen --model $MODEL [Step2 允许的参数] "prompt"
│
▼ Step 4【结果处理】
视频 = 异步:返回 task_id + status=queued(**不是失败!**) → arkcli gen get <task_id> 轮询;轮到 succeeded 自动下载到本地(local_path);要同步阻塞加 --wait(有上限,默认 10m,长视频配 --timeout 30m)
图片 = 同步:直接返回 output_url + local_path用户已经显式给出 Endpoint 时,不要先用 active profile 的模型池覆盖它:
arkcli resources resolve "$ENDPOINT" --format jsongeneration_modality 决定 image/video;image_or_video 或 unknown 时再结合
用户意图,必要时显式补 --modality。resource_region;Endpoint + 显式 API Key 且未给 Base URL 时,CLI 用该
region 派生 platform Base URL。seedream / seedance 子串猜模态。ep-* 是本次调用资源;不要忽略它后改用套餐 default,也不要把
resources resolve 的位置参数误传成模型名。若该 Endpoint 已在本轮被用户授权用于
后付费兜底,套餐额度不足时可回到这里重新做能力检查,但仍不修改 Profile/default。用户未给显式 Endpoint 时,再按 profile 列资源:
# 按目标模态列;输出 items[].id 就是可作 --model 的候选
arkcli resources list --modality video # 或 imageplatform profile → items 是推理接入点 EP(ep-xxx),每个 EP 内部绑定一个模型agent-plan / agent-plan-team → items 是套餐视觉模型名;使用对应个人/团队席位 Keycoding-plan / coding-plan-team → 无套餐内视觉模型;生成使用 platform Endpoint + 后付费 API Key。团队席位 Key 不能用于这个后付费请求is_default: true 标记的是该模态当前默认;用户没指定时优先用它invocable / required_overrides / data_plane / credential_kind。默认或可见不等于当前凭证可调用;不要为生成自动切 Profile、轮转 Key 或修改 default。$MODEL,贯穿 Step 2/3resources list 核对当前 lane 的兼容性;用户明确给了 EP 时只先 resources resolve,不要再用列表/default 覆盖它。若模型与默认不同,按 ../arkcli-shared/references/profile-defaults.md "Default 漂移检测与 promote nudge" 处理arkcli models get "$MODEL" --transform supported_params$MODEL 是模型名:拿到该模型的 supported_params 清单(每项含 name / type / support / min / max / enum / required / default / description)。MUST:Step 3 只能使用这里
support=true的参数,且取值必须落在min/max/enum范围内。 不在清单里的参数(或support=false)传了会被+gen拒绝。
MUST:
description必须逐条读,不能当注释跳过。 实测 74 个参数条目里description填充率 100%,而min/max只有 9%、enum只有 19% —— 结构化字段看着「有」的多数是空的,唯一填满的那个才是条件约束的载体。出现「仅允许 / 仅支持 / 必须 / 不支持 / 建议 / 否则」时,它是硬约束,不是提示。四类典型:
- 值域 ——
duration的min=4 max=30只给了区间,description才补上「取值为 4-30 或 -1」。-1落在区间外却合法,default=-1印证了它:default是目录自己声明的合法值,与min/max冲突时以default为准(+gen本地校验同样按此放行,无需--force)。
- 条件子集 ——
ratio的enum列了 7 个值,但「视频编辑、视频延长…仅允许 adaptive」,即某条件下enum只剩一个合法值。
- 跨参数依赖 ——
omni_reference_task_type的description给出auto/reference/edit/extend四种取值各自的ratio/duration约束;background=transparent依赖参考图的格式与数量。
- 替代方案 ——
frames的description写「请使用 duration」,即该参数不可用时该换成什么。
取值与 description 冲突、或本地校验拒绝了目录声明合法的值时,保留冲突证据(模型名、name、目录原文、CLI 报错)再决定下一步,不要静默换值或换模型。
可直接使用 Step 1 选出的模型 id:models get 会按 DisplayName 归一化到规范连字符 name。但归一化只认「点号形态 == 小写 DisplayName」这一种,不是「点号一律可用」——越界就会 not found:
doubao-seedance-2.0-fast(DisplayName 就是 Doubao-Seedance-2.0-fast)→ doubao-seedance-2-0-fastdoubao-seedream-4.5 → doubao-seedream-4-5doubao-seedream-5.0 —— 该族 DisplayName 实为 Doubao-Seedream-5.0-lite,裸族名对不上doubao-seedream-5.0-pro-260628 —— 点号 + 日期快照:DisplayName 不带日期,对不上规则:带日期快照的名字一律用连字符形态(doubao-seedream-5-0-pro-260628);点号形态只用在无日期的族名 / 变体名上。仍报 not found 时用 arkcli models search <族名> 核对规范 name,不要靠猜点号位置试。
查到模型但 supported_params 为空 / null → 该版本未配置参数目录,或上游目录当前不可解析;若 stderr 有 warn: model supported_params enrichment failed: ...,保留该告警用于排障。不要手动猜参数:+gen 会自动用内置 modality 兜底默认(video: resolution=720p / duration=5 / ratio=adaptive)填充你没指定的参数;图片任务不填 size —— 画布由 prompt 描述的宽高比与服务端默认值共同决定,硬填一个 1:1 常量会覆盖 prompt 已说清的比例,需要固定画布时显式传 --size。控制面查询失败(区别于「模型本来没配目录」)时 stderr 还会有一条 warn: ... 提示本次未做参数校验 —— 走 stderr 而非 JSON,生成失败时同样会出现。直接进 Step 3。
$MODEL 是 EP(ep-xxx):不把 EP 本身交给 models get。使用 Step 1 的权威绑定:FoundationModel 查 model_name + --version <model_version>;CustomModel 只可用 base_model_* 查 lineage 能力,不能改写真实 model_id 或调用 EP。warning/歧义时说明能力未知,不根据名称猜测支持。单个图片/视频请求不额外制造“试 Key”任务;批量候选、长视频拆段、续写链等会创建多个收费任务时,
必须在第一个 +gen 前列出总候选数、每个候选的阶段数、理论 create 总数以及阶段依赖。能用一个
原生 30 秒任务完成时,不要在模型/Endpoint 未核验前擅自拆成两个 15 秒任务。
对 Agent Plan / Agent Plan Team 的批量或多阶段视觉任务,读取当前 Profile 后执行额度快照:
arkcli usage plan --format json
arkcli usage balance --type free-quota --modality ComputerVision --page-all --format jsonresources resolve 该 EP 并按它的精确绑定重走 Step 2;否则说明缺口并停止,
不自动切 Profile、Key、default 或收费路径。控制面 resolve/list 只能证明资源元数据与上下文兼容,不能证明数据面 API Key 当前有效。没有无计费的
Key 探测时,把第一个本来就要交付的任务作为数据面准入:成功拿到 task_id/图片结果后才继续余下批次;
401/403/quota 错误按原证据停止。禁止另生成一张测试图,也禁止失败后轮转 Key 或循环试不同收费路径。
# 文生图 / 文生视频
arkcli +gen --model "$MODEL" --modality image "<prompt>" # 用户要视频时改为 video
# 带 Step 2 确认过的参数(示例:视频 1080p + 优先级 9,前提是 supported_params 列了它们)
arkcli +gen --model "$MODEL" --resolution 1080p --priority 9 "<prompt>"
# 图生图 / 图生视频 / 参考素材:--input 可重复
arkcli +gen --model "$MODEL" --input @ref.jpg "<prompt>"--input 规则、新增 --n/--priority/--wait/--timeout 见 references/arkcli-gen.mdunknown /
image_or_video 且用户意图仍不足时要求显式 --modality。--save-to <dir>);JSON 里的 local_path 是持久产物,预签名 output_url 24h 失效,优先引用 local_path。--save-to="" 关闭local_path。--open 强制打开、--no-open 强制不打开。仅对已落地本地文件生效(异步视频未 --wait 时无本地文件、不打开);多产物只打开前若干个--open:你(AI agent)调用 arkcli 时 stdout 被你接管 = 非 TTY,默认 auto 不会弹窗,用户只能看到文件路径、看不到成品。为了让用户直接看到生成的图/视频,凡是给真人出图/出视频的 +gen 与轮询到 succeeded 的 gen get,默认都加 --open(--open 无视 TTY 强制在用户桌面打开)。例外只在:用户明确说"别打开/在脚本里/批量/不要弹窗",或一次出图 >4 张批量场景 → 这时省略 --open 或显式 --no-open。| 模态 | 默认行为 | 你该怎么读结果 |
|---|---|---|
| 视频 | 异步:立即返回 task_id + status: queued | queued 不是失败。用 arkcli gen get <task_id> --open 轮询到 succeeded——这次 gen get 会顺手把产物下载到本地并回带 local_path(默认 CWD,<task-id>.mp4),--open 让成品直接在用户桌面弹出(你是 agent,非 TTY,不加就只有路径);不必再手动 curl output_url;不要因为没拿到视频就重提 +gen(会建新任务) |
视频 + --wait | 同步:阻塞到完成再返回,但超过 --timeout(默认 10m)就放弃 | arkcli +gen ... --wait --open,直接拿 output_url / local_path 并弹出成品。长视频先调大 --timeout(如 --timeout 30m):实测 seedance-2.0 近半数请求 10m 内渲不完;撞上限时返回的是「任务仍在跑 + task id」而不是失败,继续 gen get <task-id> 轮询即可,别重跑 +gen |
| 图片 | 同步:直接返回 output_url + local_path | arkcli +gen ... --open 让图片直接弹给用户看 |
⚠️ 行为变更(2.0):视频任务默认已从"自动等待完成"改为"提交即返回 task_id"。需要旧的同步阻塞行为,显式加
--wait。⚠️
--wait有上限:它最多阻塞--timeout(默认 10m),到点即返回。视频渲染常常更久(实测 seedance-2.0 近半数超过 10m),所以长视频请显式--timeout 30m。到点返回的 JSON 是type: timeout+ 带 task id 的错误——这不是生成失败,任务仍在服务端渲染;按 hint 里的gen get <task-id>继续轮询,绝不重跑+gen(会另建一个计费任务)。
gen get --format json 的 status 是对象,终态必须读 .status.phase,不是把整个 .status 与字符串比较。生成 shell 轮询脚本时必须遵守:
arkcli gen get "$TASK_ID" --save-to="" --format json 禁用自动下载,每轮只读状态。PHASE=$(printf '%s' "$RESULT" | jq -r '.status.phase // empty'),再对 succeeded / failed / cancelled 做显式分支。succeeded 时最多再执行一次带目标 --save-to 的 gen get 下载产物,然后立即 break;failed / cancelled 报告 status.message 或 error 后立即 break。queued / running 才 sleep 后继续;未知 phase 或 gen get 自身失败应停止并报错,不能当作 running 无限循环。+gen。resources list 列当前 profile 候选;模型族不确定 → 转 ../arkcli-models/SKILL.md--input @<file>(可重复)queued,用 arkcli gen get <task_id> --open 轮询;轮到 succeeded 那次会自动下载到本地(看返回的 local_path)并弹出成品,别重提--open → 你是 agent(非 TTY),不加用户只能看到路径、看不到成品;只有"别打开/脚本里/批量 >4 张"才省略或 --no-openreference_video 接受本地 @<path>(CLI 自动上传 TOS 后以预签名 URL 提交)或 https://...;本地素材上传要求账号已开通 TOS,未开通会在提交前失败并给出开通入口。ratio 逐值服从精确模型/EP 的 supported_params,不无条件强制 adaptive| 用户怎么说 | 对应 flag / 命令 |
|---|---|
| "生成完直接打开/帮我打开看看/出来就弹给我" | arkcli +gen --open(强制用系统默认程序打开;默认在交互终端已自动打开) |
| "别自动打开/不要弹窗/我在脚本里跑别开" | arkcli +gen --no-open(强制不打开) |
| "预览/别真发/只看参数/dry run/试跑/先看一下" | arkcli +gen ... --dry-run --format json;核对 steps、unresolved 和 fidelity,不要把 partial 预览当作服务端校验 |
| "不要下载/只要 URL/不要保存到本地/关闭自动下载" | 命令显式加 --save-to="";即使同时是 --dry-run 也要保留,以便预览能核对真实执行时的关闭下载意图 |
| "草稿/快速预览/越快越便宜/省钱先看" | 先区分“只看请求”和“真实生成低成本草稿”。前者用 --dry-run;后者仅在 draft 支持时加 --draft=true,不支持时说明限制,不把缩短时长冒充草稿模式 |
| "固定镜头/镜头不动/锁定相机/只拍光影变化" | 始终保留在 prompt;仅当 camera_fixed 支持该值时加 --camera-fixed=true。不支持时可用 prompt 表达视觉约束,但不能保证机械锁定,验收跨帧背景/镜头变化 |
| "不带水印/不要水印/关闭水印" | 查明支持后显式 --watermark=false;省略可能采用服务端默认值,裸 --watermark 表示 true |
| "强制执行/跳过校验/我知道不支持但想试一下" | arkcli +gen --force |
| "连贯多张/按顺序/统一风格/连续图片" | 先区分多张独立文件与一张多格图;多文件在能力支持时用 --image-count N --sequential auto,不可使用缺值的裸 --sequential |
| "我之前的任务/生成历史/任务列表/任务状态" | arkcli gen list(列出所有异步生成任务) |
| "那个任务跑完没/查进度/查状态" | arkcli gen get <task_id> |
| 命令 | 角色 |
|---|---|
arkcli resources list --modality image|video | Step 1 — 当前 profile 可用模型/EP |
arkcli resources resolve <endpoint-id> | Step 1(显式 EP) — 权威解析模态、工作流与 region |
arkcli models get <model> --transform supported_params | Step 2 — 查模型可用参数 |
arkcli +gen | Step 3 — 按可用参数生成 |
arkcli +gen --stream | 图片任务流式 NDJSON 输出 |
arkcli gen get <task-id> | Step 4 — 轮询/查询异步视频任务 |
arkcli gen list | 列出/过滤异步生成任务 |
arkcli gen delete <task-id> | 删除异步生成任务 |
not found → models get 的归一化只覆盖「点号形态 == 小写 DisplayName」,不是点号一律可用;带日期快照的名字使用连字符形态(如 doubao-seedream-5-0-pro-260628)。保留原始错误,在同一身份和范围用 arkcli models search <族名> 核对规范名称、版本与可见性;不能仅凭此错误认定模型已下线。has not activated the model / 模型未激活 → 读取 arkcli-models-activate.md,保留原模型、用户素材与生成意图,按宿主授权流程开通;成功后回到本次生成任务,不换身份、Key 或模型。已有任务 ID 时先查原任务状态,不能重复提交。若开通报 BalanceNotEnough,报告实际返回的余额/资格门槛并停止循环,不硬编码充值金额,也不把所有未激活错误解释成欠费。param_not_supported)→ 对照 Step 2 的精确版本目录与实际参数;目录显示支持但 CLI 拒绝时保留冲突证据,不擅自换调用 ID、删用户硬要求或用 --force 绕过。只有用户明确要求跳过校验时才用 --force。default:取值等于该参数的 default 却被 min/max 拒绝,就是本地校验与上游目录冲突(default 落在自己 min/max 之外时必然发生)。这类冲突不要用 --force 掩盖——先向用户报告冲突原文(模型名 / 参数名 / 目录 min·max·default / CLI 报错),由用户决定是修 CLI 还是显式要求 --force。用 --force 绕过会把「CLI 有 bug」永久伪装成「参数已提交」,下一个用户会再撞一次。ContentRiskBlocked / *SensitiveContentDetected / 命中敏感 / 版权)→ 不是参数问题、--force 也绕不过;调整 prompt / 输入素材里的敏感内容后重试。要结构化的拦截原因 + 修复指引,转 ../arkcli-doctor/SKILL.md 的 arkcli doctor error <code>(生视频拦截 5 个 subtype 全覆盖)../arkcli-auth/SKILL.mdsupported_params 详解+gen 全参数 + 多模态 + 异步语义© volcengine, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 5 other files (references) in skills/arkcli-gen of volcengine/ark-cli.
Open the folder on GitHubat commit fb5b7be
Arkcli Gen next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Arkcli Gen this skillvolcengine/ark-cli | 140 | — | ~3.9k | Automated safety check: Pass | Apache-2.0 | |
| Gh Stackremotion-dev/remotion | 62k | 6 repos | ~2.4k | Automated safety check: Pass | Custom licence | |
| HyperFrames Media Useheygen-com/hyperframes | 58k | 2 repos | ~2.1k | Automated safety check: Pass | Apache-2.0 | |
| Guizang Social Cardsop7418/guizang-social-card-skill | 7.4k | 1 repos | ~7.8k | Automated safety check: Pass | AGPL-3.0 | |
| Weekly Changelog Videoheygen-com/hyperframes | 58k | — | ~3.3k | Automated safety check: Pass | Apache-2.0 | |
| Anthropic Brand Stylinganthropics/skills | 180k | 29 repos | ~559 | Automated safety check: Pass | Apache-2.0 |
remotion-dev/remotion
Manages stacked PRs and splits multi-part work into reviewable branches with gh-stack.
heygen-com/hyperframes
Finds, generates and edits media for HyperFrames video projects: music, sound effects, images, icons, logos, voiceovers, captions and color grades.
op7418/guizang-social-card-skill
Produces social card sets for Xiaohongshu and WeChat: carousels, Live Photo motion cards and puzzle layouts, and WeChat cover pairs, rendered from single-file HTML.
heygen-com/hyperframes
Turns a weekly changelog markdown file into a branded HyperFrames video with voiceover, animated mock-UI scenes and captions, using fonts, background and scripts bundled in the skill.
anthropics/skills
Applies Anthropic's brand colors and fonts to artifacts such as PowerPoint slides, using fixed hex values for text and accents, Poppins headings and Lora body text.
harry0703/MoneyPrinterTurbo
Installs and runs MoneyPrinterTurbo to turn a topic or script into a finished short video with voice-over, subtitles, stock footage and music.
volcengine/ark-cli
Inspect or invoke locally registered ArkCLI actions when product commands cannot cover a task.
volcengine/ark-cli
arkcli +code-example:为指定基础模型生成多语言(Python / Go / Java / Node / curl)调用示例代码并写入本地文件。数据源是火山方舟 OpenTOP OpenGetSampleCode。当用户需要拿某个基础模型的 SDK / curl 调用示例、保存为本地接入模板时使用。反触发:TTS/ASR/语音模型没有 arkcli…
volcengine/ark-cli
arkcli 本地配置管理。处理 profile 配置归因、update.mode 的 automatic/disabled 策略、config reset 与历史 yaml 排障;profile 类操作优先使用 arkcli profile <subcmd。
volcengine/ark-cli
arkcli 自定义模型仓库管理:从 TOS 导入自定义模型、查询/筛选自定义模型、查看详情、改名、删除、查询可用量化模式、量化已就绪的模型。任何提到自定义模型 ID(cm-)的管理、部署准备,或要求用 cm- 直接对话/推理/试效果的边界判断,都必须使用本 skill。注意:查询火山公共基础模型(doubao 等 foundation models)走 arkcli-models;本…
volcengine/ark-cli
arkcli +deploy:普通创建推理接入点(Endpoint)的统一首选入口。用户说『创建/新建/create 一个 endpoint/接入点』或『部署/上线/deploy 某模型』时优先走这里;但脚本化 / CI / 无护栏 / 原始 raw CRUD 创建是唯一例外,必须改走 arkcli-infer-endpoint,不能由本 skill…
volcengine/ark-cli
检索、读取与总结方舟官方文档。用户给出 ark.volcengine.com 文档 URL 或 /docs/ 路径、要求读链接、官方说明、API 契约或必填字段,询问 CC Switch 等第三方客户端的方舟图形配置流程,以及官方网页读取失败时使用。不用于业务调用、资源操作、CLI 帮助或通用知识。
Categories
火山方舟 Ark 图片/视频生成入口:支持 profile 默认资源与临时 API Key/Base URL/Endpoint;显式 Endpoint 不受当前 plan profile 误导。图片同步返回,视频异步轮询。. Arkcli Gen is an agent skill from volcengine/ark-cli.
Arkcli Gen fits situations like: media & Creative work in your project.
Run `npx skills add volcengine/ark-cli --skill arkcli-gen -a claude-code`. Or copy the skill folder (skills/arkcli-gen in volcengine/ark-cli) into .claude/skills/arkcli-gen in your project. Claude Code loads it when a task matches its description.
Run `npx skills add volcengine/ark-cli --skill arkcli-gen -a codex`. Or copy the skill folder (skills/arkcli-gen in volcengine/ark-cli) into .agents/skills/arkcli-gen in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add volcengine/ark-cli --skill arkcli-gen -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/arkcli-gen, .gemini/skills/arkcli-gen, .github/skills/arkcli-gen and .opencode/skills/arkcli-gen in your project.
Going by SKILL.md and its folder, Arkcli Gen needs the command-line tools its instructions call (jq).
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Arkcli Gen is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.9k tokens (SKILL.md is roughly 16k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 14k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Arkcli Gen: Gh Stack (remotion-dev/remotion, 62k stars), HyperFrames Media Use (heygen-com/hyperframes, 58k stars), Guizang Social Cards (op7418/guizang-social-card-skill, 7.4k stars) and Weekly Changelog Video (heygen-com/hyperframes, 58k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
volcengine (a GitHub organization) maintains it in volcengine/ark-cli, which has 140 GitHub stars. The repository holds 25 skills in this directory. The repository was last updated on September 29, 2026.
Source: volcengine/ark-cli on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.