@genspark/cli 1.15.4 → 1.15.6
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +31 -11
- package/dist/acp-serve.d.ts +2 -2
- package/dist/acp-serve.d.ts.map +1 -1
- package/dist/acp-serve.js +7 -2
- package/dist/acp-serve.js.map +1 -1
- package/dist/client.d.ts +5 -0
- package/dist/client.d.ts.map +1 -1
- package/dist/client.js +23 -12
- package/dist/client.js.map +1 -1
- package/dist/commands/aidrive-contract.generated.d.ts.map +1 -1
- package/dist/commands/aidrive-contract.generated.js +21 -2
- package/dist/commands/aidrive-contract.generated.js.map +1 -1
- package/dist/commands/task-status-follow.d.ts +9 -0
- package/dist/commands/task-status-follow.d.ts.map +1 -1
- package/dist/commands/task-status-follow.js +24 -4
- package/dist/commands/task-status-follow.js.map +1 -1
- package/dist/confirm.d.ts.map +1 -1
- package/dist/confirm.js +18 -1
- package/dist/confirm.js.map +1 -1
- package/dist/humanApproval.d.ts +48 -0
- package/dist/humanApproval.d.ts.map +1 -0
- package/dist/humanApproval.js +86 -0
- package/dist/humanApproval.js.map +1 -0
- package/dist/index.js +88 -21
- package/dist/index.js.map +1 -1
- package/dist/mediaSummary.d.ts.map +1 -1
- package/dist/mediaSummary.js +4 -1
- package/dist/mediaSummary.js.map +1 -1
- package/dist/task-shim.d.ts +11 -2
- package/dist/task-shim.d.ts.map +1 -1
- package/dist/task-shim.js +21 -2
- package/dist/task-shim.js.map +1 -1
- package/docs/skills.md +1 -1
- package/gated-skills/gsk-issue/SKILL.md +7 -0
- package/gated-skills/gsk-m365-apps/SKILL.md +197 -0
- package/gated-skills/gsk-m365-apps/references/add-genspark-ai.md +399 -0
- package/gated-skills/gsk-m365-apps/scripts/install-genspark-connector.mjs +548 -0
- package/gated-skills/gsk-m365-apps/scripts/sync-microsoft-guidance.mjs +557 -0
- package/package.json +1 -1
- package/skills/gsk-aidrive/SKILL.md +14 -2
- package/skills/gsk-aidrive/command-reference.md +13 -3
- package/skills/gsk-image-generation/SKILL.md +6 -3
- package/skills/gsk-phone-call/SKILL.md +33 -13
- package/skills/gsk-second-brain/SKILL.md +2 -2
- package/skills/gsk-slack/SKILL.md +4 -1
- package/skills/gsk-team-brain/SKILL.md +89 -22
- package/skills/gsk-video-generation/SKILL.md +3 -3
|
@@ -29,9 +29,12 @@ gsk img [options]
|
|
|
29
29
|
|
|
30
30
|
| Flag | Required | Description |
|
|
31
31
|
|------|----------|-------------|
|
|
32
|
-
| `-m`, `--model` | No | If this parameter is omitted, nano-banana-2-flash-lite is used. Pick ONLY a model described below; never pick an undescribed model name unless the user explicitly asks for it by name. gpt-image-2: GPT Image 2 - Latest OpenAI image model. Superior text rendering, editing, and face preservation. Supports text-to-image and image editing with multi-image input. gpt-image-2.5: GPT Image 2.5 - OpenAI's newest image model and the default for high-fidelity generations. Edits change only what you ask, keeping subject and composition; natural lighting, rich textures. Text-to-image and editing with multi-image input, 1k-4k output. Pick a tier with `tier` (flare = fast default arm, sunburst = precision edits; same price). Quality rungs low/medium/high/xhigh/max (auto = high): xhigh and max bill extra credits on non-Pro plans, so only use them when the user explicitly asks for that quality. fal-ai/bytedance/seedream/v5/lite: Bytedance Seedream v5 Lite model for text-to-image, image-to-image (single and multiple image editing) generation with native 2K resolution and excellent text layout.fal-ai/bytedance/seedream/v5/pro: Bytedance Seedream 5.0 Pro - flagship photorealistic model for text-to-image and multi-image editing (up to 10 reference images). High-density text/infographic layout, region-precise edits, native text in 14 languages, up to 2K output. xai/grok-imagine-image: xAI Grok Imagine Image 2.0 - top-ranked text-to-image and image editing (up to 3 reference images). Strong photorealism and prompt adherence, up to 2K output. qwen-image-3: Qwen Image 3 - Alibaba's latest image model for text-to-image and reference-image editing (1-3 reference images). Superior Chinese & English text rendering, precise prompt adherence, identity-preserving edits. Best for posters, bilingual typography, and instruction-based edits. fal-ai/flux-3: BFL FLUX 3 Image - newest Black Forest Labs model with native 2K/4K output, precise layout control and improved text rendering. Text-to-image and editing with up to 10 reference images. Ideal for posters, ads, infographics and layout-sensitive realistic images. fal-ai/z-image/turbo: Z-Image Turbo model optimized for speed without sacrificing quality. Supports both text-to-image and image-to-image modes. Perfect for quick iterations, bulk generation, style transfer, and cost-effective projects.bbox-segment: A model for extracting subjects from images based on bbox region.fal-bria-rmbg: A model specialized for remove background from image. Should prioritize this tool if the user wants to remove background from image. fal-ai/recraft-clarity-upscale: A model specialized for upscale image. fal-ai/image-editing/text-removal: A specialized model for removing text and watermarks from images. Automatically detects and removes text while preserving the background and other visual elements. Use when users want to clean images, remove watermarks, or eliminate text overlays. flux-pro/outpaint: A model specialized for expand image to a specific aspect ratio. nano-banana-2-flash-lite: [GEMINI NANO BANANA 2 FLASH LITE] Fast, lightweight Gemini image model for text-to-image and image editing with multi-image input. Native 1K output. Best for speed-sensitive, high-volume workflows. nano-banana-2.1: [GEMINI NANO BANANA 2.1] Google's newest Gemini image model: a quality step up from Nano Banana 2 for text-to-image and editing with up to 14 reference images, 1K/2K/4K output and a selectable thinking depth (thinking_level). Pick it when the user asks for it by name or wants the best Gemini result. (string, one of: gpt-image-2, gpt-image-2.5, fal-ai/bytedance/seedream/v5/lite, fal-ai/bytedance/seedream/v5/pro, xai/grok-imagine-image, qwen-image-3, fal-ai/flux-3, fal-ai/z-image/turbo, bbox-segment, fal-bria-rmbg, fal-ai/recraft-clarity-upscale, fal-ai/image-editing/text-removal, flux-pro/outpaint, nano-banana-2-flash-lite, nano-banana-2.1) |
|
|
33
|
-
| `--thinking_level` | No | nano-banana-2.1 only (ignored by other models): how much the model thinks before rendering.
|
|
34
|
-
| `--tier` | No | gpt-image-2.5 only (ignored by other models): which arm of the model runs. 'flare' (default when omitted) is the fast arm for most work; 'sunburst' trades speed for tighter control across repeated edits. Same price. (string, one of: flare, sunburst) |
|
|
32
|
+
| `-m`, `--model` | No | If this parameter is omitted, nano-banana-2-flash-lite is used. Pick ONLY a model described below; never pick an undescribed model name unless the user explicitly asks for it by name. gpt-image-2: GPT Image 2 - Latest OpenAI image model. Superior text rendering, editing, and face preservation. Supports text-to-image and image editing with multi-image input. gpt-image-2.5: GPT Image 2.5 - OpenAI's newest image model and the default for high-fidelity generations. Edits change only what you ask, keeping subject and composition; natural lighting, rich textures. Text-to-image and editing with multi-image input, 1k-4k output. Pick a tier with `tier` (flare = fast default arm, sunburst = precision edits; same price). Quality rungs low/medium/high/xhigh/max (auto = high): xhigh and max bill extra credits on non-Pro plans, so only use them when the user explicitly asks for that quality. gpt-image-2.6: GPT Image 2.6 - OpenAI's newest image model (2026-11-04), same contract as 2.5 (edits change only what you ask; text-to-image and editing with multi-image input, 1k-4k) plus a planning-depth knob: `thinking_level` low/medium/high (default high, no extra fee; low = faster focused edits, high = dense text and complex layouts). Pick a tier with `tier` (flare = fast default arm, sunburst = detailed work and precision edits; same price). Quality rungs low/medium/high/xhigh/max (auto = high): low and medium cost 3-4x 2.5's rungs, so low is not a cheap draft; xhigh and max bill extra credits on non-Pro plans, so only use them when the user explicitly asks. fal-ai/bytedance/seedream/v5/lite: Bytedance Seedream v5 Lite model for text-to-image, image-to-image (single and multiple image editing) generation with native 2K resolution and excellent text layout.fal-ai/bytedance/seedream/v5/pro: Bytedance Seedream 5.0 Pro - flagship photorealistic model for text-to-image and multi-image editing (up to 10 reference images). High-density text/infographic layout, region-precise edits, native text in 14 languages, up to 2K output. xai/grok-imagine-image: xAI Grok Imagine Image 2.0 - top-ranked text-to-image and image editing (up to 3 reference images). Strong photorealism and prompt adherence, up to 2K output. qwen-image-3: Qwen Image 3 - Alibaba's latest image model for text-to-image and reference-image editing (1-3 reference images). Superior Chinese & English text rendering, precise prompt adherence, identity-preserving edits. Best for posters, bilingual typography, and instruction-based edits. fal-ai/flux-3: BFL FLUX 3 Image - newest Black Forest Labs model with native 2K/4K output, precise layout control and improved text rendering. Text-to-image and editing with up to 10 reference images. Ideal for posters, ads, infographics and layout-sensitive realistic images. fal-ai/z-image/turbo: Z-Image Turbo model optimized for speed without sacrificing quality. Supports both text-to-image and image-to-image modes. Perfect for quick iterations, bulk generation, style transfer, and cost-effective projects.bbox-segment: A model for extracting subjects from images based on bbox region.fal-bria-rmbg: A model specialized for remove background from image. Should prioritize this tool if the user wants to remove background from image. fal-ai/recraft-clarity-upscale: A model specialized for upscale image. fal-ai/image-editing/text-removal: A specialized model for removing text and watermarks from images. Automatically detects and removes text while preserving the background and other visual elements. Use when users want to clean images, remove watermarks, or eliminate text overlays. flux-pro/outpaint: A model specialized for expand image to a specific aspect ratio. nano-banana-2-flash-lite: [GEMINI NANO BANANA 2 FLASH LITE] Fast, lightweight Gemini image model for text-to-image and image editing with multi-image input. Native 1K output. Best for speed-sensitive, high-volume workflows. nano-banana-2.1: [GEMINI NANO BANANA 2.1] Google's newest Gemini image model: a quality step up from Nano Banana 2 for text-to-image and editing with up to 14 reference images, 1K/2K/4K output and a selectable thinking depth (thinking_level). Pick it when the user asks for it by name or wants the best Gemini result. bytedance/seedream-5.5: [BYTEDANCE SEEDREAM 5.5] Text-to-image and edits (up to 10 references), 1K/2K, dense text; natively outputs editable layers (output_layers / max_layers) and transparent PNGs (background). (string, one of: gpt-image-2, gpt-image-2.5, gpt-image-2.6, fal-ai/bytedance/seedream/v5/lite, fal-ai/bytedance/seedream/v5/pro, xai/grok-imagine-image, qwen-image-3, fal-ai/flux-3, fal-ai/z-image/turbo, bbox-segment, fal-bria-rmbg, fal-ai/recraft-clarity-upscale, fal-ai/image-editing/text-removal, flux-pro/outpaint, nano-banana-2-flash-lite, nano-banana-2.1, bytedance/seedream-5.5) |
|
|
33
|
+
| `--thinking_level` | No | nano-banana-2.1 and gpt-image-2.6 only (ignored by other models): how much the model thinks or plans before rendering. nano-banana-2.1 takes minimal/medium/high (default medium; minimal is fastest); gpt-image-2.6 takes low/medium/high (default high; low is faster for focused edits). Omit it for the default. Same price. (string, one of: minimal, low, medium, high) |
|
|
34
|
+
| `--tier` | No | gpt-image-2.5 and gpt-image-2.6 only (ignored by other models): which arm of the model runs. 'flare' (default when omitted) is the fast arm for most work; 'sunburst' trades speed for tighter control across repeated edits and detailed work. Same price. (string, one of: flare, sunburst) |
|
|
35
|
+
| `--output_layers` | No | Models whose info lists layers: return editable layers (base + one transparent PNG per element) instead of one flat image. Bills per layer. Only when the user asks for layers. Default false. (boolean) |
|
|
36
|
+
| `--max_layers` | No | Layered output only: cap on outputs (base + layers), 2-100; 30 is the default, not a limit. (integer) |
|
|
37
|
+
| `--background` | No | Models whose info lists it: 'transparent' (PNG with alpha), 'opaque' or 'auto' (default). Flat images only. A single transparent reference keeps its alpha unless this is set; pass 'auto' to let the model fill a requested background. (string, one of: auto, transparent, opaque) |
|
|
35
38
|
| `<query>` (positional) | No | Detailed description of the image to generate. The prompt should be in English.However, any specific text that should appear within the image does not need to be translated.(e.g., 'A futuristic cityscape with flying cars and neon lights at night'). Omit when passing the prompt by `query_file` instead. (string) |
|
|
36
39
|
| `--query_file` | No | Optional. Repo-relative path to a UTF-8 text file containing the prompt, used verbatim in place of `query` (lets you pass a long prompt by path instead of inlining it). Requires `repo_id`. When set, `query` may be omitted. (string) |
|
|
37
40
|
| `--bbox` | No | BBox of the target object, for bbox segmentation modelFormat: [x1, y1, x2, y2], x1 and y1 are the top-left corner coordinates, x2 and y2 are the bottom-right corner coordinates. (array) |
|
|
@@ -5,11 +5,12 @@ description: Make a real AI phone call on the user's behalf. Validates prerequis
|
|
|
5
5
|
(membership, phone setup, credits), resolves the contact, then dials and returns
|
|
6
6
|
status, real audio duration, dial attempts, summary, and transcript. After every
|
|
7
7
|
call, present the outcome, disclose every dial attempt, point to call_log or call_detail,
|
|
8
|
-
and show short transcripts inline. Never fabricate a phone number
|
|
9
|
-
|
|
10
|
-
a call. On code=setup_incomplete, show
|
|
11
|
-
|
|
12
|
-
|
|
8
|
+
and show short transcripts inline. Never fabricate a phone number; phone_number
|
|
9
|
+
must have verified lineage, and contact_ref re-dials a call-log contact or self.
|
|
10
|
+
A dry run validates only and never places a call. On code=setup_incomplete, show
|
|
11
|
+
setup_url, wait for setup, and use phone_call_setup_status before retrying. For
|
|
12
|
+
sending a real text message instead, use the SMS sibling `gsk telephony sms send`
|
|
13
|
+
— same setup, lineage, and disclosure rules.
|
|
13
14
|
metadata:
|
|
14
15
|
category: phone-call
|
|
15
16
|
generated_by: genspark_tool_cli/generate_skills.py
|
|
@@ -23,7 +24,7 @@ metadata:
|
|
|
23
24
|
|
|
24
25
|
**PREREQUISITE:** Read `../gsk-shared/SKILL.md` for auth, global flags, and security rules.
|
|
25
26
|
|
|
26
|
-
Make a real AI phone call on the user's behalf. Validates prerequisites (membership, phone setup, credits), resolves the contact, then dials and returns status, real audio duration, dial attempts, summary, and transcript. After every call, present the outcome, disclose every dial attempt, point to call_log or call_detail, and show short transcripts inline. Never fabricate a phone number
|
|
27
|
+
Make a real AI phone call on the user's behalf. Validates prerequisites (membership, phone setup, credits), resolves the contact, then dials and returns status, real audio duration, dial attempts, summary, and transcript. After every call, present the outcome, disclose every dial attempt, point to call_log or call_detail, and show short transcripts inline. Never fabricate a phone number; phone_number must have verified lineage, and contact_ref re-dials a call-log contact or self. A dry run validates only and never places a call. On code=setup_incomplete, show setup_url, wait for setup, and use phone_call_setup_status before retrying. For sending a real text message instead, use the SMS sibling `gsk telephony sms send` — same setup, lineage, and disclosure rules.
|
|
27
28
|
|
|
28
29
|
## Usage
|
|
29
30
|
|
|
@@ -34,8 +35,9 @@ gsk telephony call dial [options]
|
|
|
34
35
|
## SMS sibling: `gsk telephony sms send`
|
|
35
36
|
|
|
36
37
|
Send a real SMS on the user's behalf. The message body is sent VERBATIM — pass exactly the text the user approved, never compose or embellish it. Validates prerequisites (membership, phone setup, credits), destination number lineage, and the per-country segment budget (South Korea: single segment only, 70 chars for Korean text; Japan: max 5 segments), then sends and returns the delivery handle (sid), segments, and credits charged. Over-budget messages are rejected, never truncated.
|
|
37
|
-
Destination numbers must have verifiable lineage (user-provided, prior tool results, or a calllog: contact_ref) — never fabricate a number.
|
|
38
|
+
Destination numbers must have verifiable lineage (user-provided, prior tool results, or a calllog: contact_ref) — never fabricate a number. The number must be E.164 ('+' + country calling code + national number); a number without its country calling code is refused (code invalid_phone_number).
|
|
38
39
|
Use dry_run=true first when unsure: it validates and prices (encoding, segment count, credits) without sending.
|
|
40
|
+
Read the outcome from the envelope's session_state, as for phone_call — the envelope status is ok even when nothing was sent: a refusal sets session_state.sms_precheck to login_required, membership_required, setup_incomplete or insufficient_credits (the message text carries the setup or pricing URL); a dry run or a send sets session_state.sms_result (to, country, encoding, segments, max_segments, estimated_credits; a dry run adds assembled_message, a send adds sms_id, status, credits_charged). Branch on these fields, never on the message text.
|
|
39
41
|
MANDATORY after every send: report to the user the recipient, delivery status, segment count, and credits charged.
|
|
40
42
|
|
|
41
43
|
```bash
|
|
@@ -65,8 +67,15 @@ phone task; the sections above hold the command details.
|
|
|
65
67
|
2. **Check readiness.** Run `gsk telephony setup-status` whenever the ask is
|
|
66
68
|
about calling. Not ready → hand the user the returned `setup_url` and stop;
|
|
67
69
|
never dial around a gate. The number and the recipient name come from the user, prior tool results
|
|
68
|
-
or research (a maps search
|
|
69
|
-
invented.
|
|
70
|
+
or research (a maps search lists the phone number) — never
|
|
71
|
+
invented. **The number must be E.164**: `+` + country calling code +
|
|
72
|
+
national number (`+<country calling code><national number>`). Every
|
|
73
|
+
number must carry its country calling code; one without it is refused
|
|
74
|
+
with `code=invalid_phone_number` and the country is never guessed for
|
|
75
|
+
you. The refusal lists the clues the server has (where the request came
|
|
76
|
+
from — for you in a sandbox that is the sandbox, not the user — and the
|
|
77
|
+
country of the user's own verified number); use them to ASK the user
|
|
78
|
+
which country the number belongs to, then retry with the full `+` form.
|
|
70
79
|
3. **Write a self-sufficient `purpose`.** One string carrying the goal, every
|
|
71
80
|
fact the callee may need, every question that must be answered, what to
|
|
72
81
|
accept and what to decline, and the call language. Nothing can be added
|
|
@@ -88,8 +97,12 @@ phone task; the sections above hold the command details.
|
|
|
88
97
|
called too often in the last 24 hours is refused. Never write a purpose
|
|
89
98
|
that argues around any of this.
|
|
90
99
|
4. **Dial once, asynchronously, then poll.**
|
|
91
|
-
`gsk telephony call dial <recipient> -
|
|
92
|
-
|
|
100
|
+
`gsk telephony call dial <recipient> -n <phone number> -p <purpose> --async`
|
|
101
|
+
(or `-r <calllog: contact_ref | self>` to re-dial by reference)
|
|
102
|
+
returns at once with `project_id`. The dispatch reply carries no dial
|
|
103
|
+
state: its `call_status` is `preparing` — the purpose review has not run
|
|
104
|
+
and nothing has been dialed; only the status poll tells you whether the
|
|
105
|
+
call was placed. Poll `gsk telephony call status
|
|
93
106
|
<project_id>` until the status is terminal. A call can run several minutes
|
|
94
107
|
through IVR and hold; a long call still in progress is not a stuck call. If
|
|
95
108
|
a blocking dial timed out, the call is still live: take the project id from
|
|
@@ -160,6 +173,12 @@ phone task; the sections above hold the command details.
|
|
|
160
173
|
transcript; `gsk telephony call recording <project_id> ./call.wav` saves
|
|
161
174
|
the audio file (see "Getting the recording file" above). Redirect stdout
|
|
162
175
|
to a file and parse the file: long transcripts truncate in context.
|
|
176
|
+
The persisted record can lag `status` by a few seconds after it turns
|
|
177
|
+
terminal: a `detail` with `"recording_pending": true` has no recording
|
|
178
|
+
links yet, so retry it for the recording. A call that did not connect
|
|
179
|
+
has no recording at all; its `detail` carries the same diagnostics as
|
|
180
|
+
`status` (`sip_response_code`, `error_code`, `error_message`,
|
|
181
|
+
`dial_rejection_reason`, `dial_attempts`).
|
|
163
182
|
6. **Judge every `completed` call from the transcript, not from the
|
|
164
183
|
status.** `completed` includes voicemail, refusals and "we will call you
|
|
165
184
|
back". Goal met → the task ends; do not dial again to "confirm".
|
|
@@ -183,8 +202,9 @@ call them one at a time and stop at the first acceptable result; for a
|
|
|
183
202
|
|
|
184
203
|
| Flag | Required | Description |
|
|
185
204
|
|------|----------|-------------|
|
|
186
|
-
| `<recipient>` (positional) | Yes | The name of the person or business to call.
|
|
187
|
-
| `-
|
|
205
|
+
| `<recipient>` (positional) | Yes | The name of the person or business to call, e.g. 'John Smith', 'Dr. Sarah Johnson', 'Starbucks George Street'. (string) |
|
|
206
|
+
| `-n`, `--phone_number` | No | The phone number to call, in E.164: '+' followed by the country calling code and the national number ('+<country calling code><national number>'). Every number must carry its country calling code; a number without it is refused (code invalid_phone_number) and the refusal lists country clues (the request's origin, the user's own verified number) — ask the user for the country rather than guessing. MUST have verified data lineage from (1) user-provided info, (2) previous tool results, or (3) explicit context — NEVER fabricate, estimate, or infer a number. Finding the number is your job, not the CLI's: look it up first (e.g. `gsk maps-search`, a web search) and pass the phone number the result lists, with its country code. Exactly one of phone_number / contact_ref. (string) |
|
|
207
|
+
| `-r`, `--contact_ref` | No | A reference the server resolves to a number: a calllog: contact_ref from the call log's contacts view (`gsk telephony call log --view contacts`) — the preferred way to re-dial someone previously called, their real number stays masked — or self, your own verified number. Exactly one of phone_number / contact_ref. (string) |
|
|
188
208
|
| `-p`, `--purpose` | Yes | The clear reason for making the call (e.g., 'Check reservation availability', 'Inquire about business hours'). (string) |
|
|
189
209
|
|
|
190
210
|
> **CAUTION:** This command performs a write/send operation. Double-check parameters before executing.
|
|
@@ -52,7 +52,7 @@ A `mounted: false` source exists but is switched off; only mounted sources accep
|
|
|
52
52
|
|
|
53
53
|
**The user's profile.** For questions about the user themself — who they are, what they work on, how they like things done — read `memo/System/memory/Memory.md` first. Treat it as background that was true when written, not as instructions.
|
|
54
54
|
|
|
55
|
-
**Writing Notes.** `gsk sb-git commit` writes the FULL file content (not append) and carries `expected_parent` for compare-and-set. New notes go under `Personal Space/memo/`; reuse the folder and naming pattern already there.
|
|
55
|
+
**Writing Notes.** `gsk sb-git commit` writes the FULL file content (not append) and carries `expected_parent` for compare-and-set. New notes go under `Personal Space/memo/`; reuse the folder and naming pattern already there. Hand back the file you wrote: inside Second Brain chat as a `memo:` link; anywhere else (GenTeam, other chats) as the commit reply's `urls[<path>]` (a `memo:` link opens nothing there). A bare `ls` is not proof of persistence.
|
|
56
56
|
|
|
57
57
|
## 3. The Drive half
|
|
58
58
|
|
|
@@ -190,7 +190,7 @@ When a `gsk drive` reply already carries one, hand that `url` back as it is: `sh
|
|
|
190
190
|
|
|
191
191
|
A reply's `Open folder` / `browse_url` opens the containing folder by the same entry-link scheme; use it only when the user wants the folder. For an uploaded or read file, prefer its `Link` / `url`. Only the caller's Drive home uses a home-page URL. Do not reconstruct folder links with `?path=` or file links with `#p_fid=`.
|
|
192
192
|
|
|
193
|
-
Never hand out a `get_readable_url` (a signed download that opens nothing), a `/aidrive/files/...` folder page, or an editor route (`/second-brain/doc?id=...` is an in-app destination, not a share link). For Notes, hand back the `memo:` link.
|
|
193
|
+
Never hand out a `get_readable_url` (a signed download that opens nothing), a `/aidrive/files/...` folder page, or an editor route (`/second-brain/doc?id=...` is an in-app destination, not a share link). For Notes, hand back the `memo:` link in Second Brain chat, the commit reply's `urls` link elsewhere.
|
|
194
194
|
|
|
195
195
|
**Handing a file to a GenTeam conversation.** Choose one delivery for each result, within the current turn's permissions:
|
|
196
196
|
|
|
@@ -43,7 +43,10 @@ gsk slack [options]
|
|
|
43
43
|
| `--question` | No | [search] A specific question to answer based on the search results. (string) |
|
|
44
44
|
| `--channel` | No | [read] Conversation ID (channel 'C…', private group 'G…', or DM 'D…') or a channel name (e.g. '#general' or 'general'). IDs come from the lookup action. \| [react] Conversation ID containing the message (channel 'C…', private group 'G…', or DM 'D…'). \| [invite] Conversation ID of the channel to invite into (public 'C…' or private group 'G…'). \| [upload] Target conversation ID (channel 'C…', private group 'G…', or DM 'D…') the file is shared into. \| [update] Conversation ID containing the message (channel 'C…', private group 'G…', or DM 'D…'). \| [delete] Conversation ID containing the message (channel 'C…', private group 'G…', or DM 'D…'). \| [permalink] Conversation ID containing the message (channel 'C…', private group 'G…', or DM 'D…'). (string) |
|
|
45
45
|
| `--limit` | No | [read] Maximum messages to return (default 20, max 100). \| [lookup] Maximum number of results to return. Default: 50 (integer) |
|
|
46
|
-
| `--cursor` | No | [read] Pagination cursor from a previous call's next_cursor to fetch older
|
|
46
|
+
| `--cursor` | No | [read] Pagination cursor from a previous call's next_cursor to fetch the next page (older history or newer replies). \| [lookup] Unfiltered channels only: use lookup_type='channels' without search_query. Pass next_cursor from the previous page in the same workspace. (string) |
|
|
47
|
+
| `--oldest` | No | [read] Read messages after this Unix timestamp (seconds, optionally fractional). (string) |
|
|
48
|
+
| `--latest` | No | [read] Read messages before this Unix timestamp (seconds, optionally fractional). (string) |
|
|
49
|
+
| `--inclusive` | No | [read] Include messages at oldest/latest boundaries. Default false; ignored when neither bound is set. (boolean) |
|
|
47
50
|
| `--file_id` | No | [download_file] Slack file id (starts with 'F'), from a message's files list (read action) or a search_files hit. (string) |
|
|
48
51
|
| `--lookup_type` | No | [lookup] What to look up: 'users' for team members, 'channels' for channels/conversations, 'all' for both, 'workspaces' for the Slack workspaces linked to this account (their team_id feeds the 'workspace' parameter). (string, one of: users, channels, all, workspaces) |
|
|
49
52
|
| `--search_query` | No | [lookup] Optional search query to filter results by name. Case-insensitive partial match on name, display name, or real name. (string) |
|
|
@@ -1,8 +1,8 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: gsk-team-brain
|
|
3
3
|
version: 1.0.0
|
|
4
|
-
description: Query
|
|
5
|
-
|
|
4
|
+
description: Query Team Brains or public product help and read cited evidence; create
|
|
5
|
+
and publish a new Team Brain.
|
|
6
6
|
metadata:
|
|
7
7
|
category: general
|
|
8
8
|
generated_by: genspark_tool_cli/generate_skills.py
|
|
@@ -16,7 +16,7 @@ metadata:
|
|
|
16
16
|
|
|
17
17
|
**PREREQUISITE:** Read `../gsk-shared/SKILL.md` for auth, global flags, and security rules.
|
|
18
18
|
|
|
19
|
-
Query
|
|
19
|
+
Query Team Brains or public product help and read cited evidence; create and publish a new Team Brain.
|
|
20
20
|
|
|
21
21
|
## Usage
|
|
22
22
|
|
|
@@ -54,12 +54,22 @@ coverage to choose relevant Brains; unrelated questions need no Brain call.
|
|
|
54
54
|
|
|
55
55
|
- Query with `gsk --output json team-brain ask -Q '<question>' --mode retrieval`.
|
|
56
56
|
Include the conversation context needed to understand the question.
|
|
57
|
-
-
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
|
|
57
|
+
- Retrieval is fast by default: one bounded hybrid search per query, without an
|
|
58
|
+
inner retrieval LLM. Add `--queries` with the question in the Brain's other
|
|
59
|
+
language and the specific entity, product or company names involved; a
|
|
60
|
+
Chinese question over English pages, or the reverse, otherwise ranks poorly.
|
|
61
|
+
- The response holds original windows for the top files in retrieval.evidence,
|
|
62
|
+
more ranked pages in navigation.pointers (path, line, snippet) and index lines
|
|
63
|
+
in navigation.directory. Pointers and directory entries are not evidence: read
|
|
64
|
+
the ones you need in one call with
|
|
65
|
+
`gsk --output json team-brain read-source --brain <brain_id> --pages <path> <path> --version <head_sha>`.
|
|
66
|
+
- To enumerate a topic (all deals in a market, decisions in a month), grep a
|
|
67
|
+
folder or the index page a directory entry names:
|
|
68
|
+
`gsk --output json team-brain grep --brain <brain_id> --pattern 'NTT|KDDI' --folder <folder or index.md>`,
|
|
69
|
+
then read the matching pages together. Check relevance, dates and conditions
|
|
70
|
+
in the originals. Use `--retrieval_strategy deep` for progressive
|
|
71
|
+
investigation by the server when evidence stays insufficient. Do not repeat
|
|
72
|
+
the identical query to recover saved evidence.
|
|
63
73
|
- Without --brain, ask queries all currently available Brains. A catalog
|
|
64
74
|
brain_id passed as --brain only narrows that set.
|
|
65
75
|
Human callers use their available organization catalog. For GenTeam agent calls,
|
|
@@ -99,11 +109,41 @@ coverage to choose relevant Brains; unrelated questions need no Brain call.
|
|
|
99
109
|
|
|
100
110
|
## Editing
|
|
101
111
|
|
|
102
|
-
For a user-requested edit, read `gsk team-brain describe --topic edit
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
112
|
+
For a user-requested edit, read `gsk team-brain describe --topic edit`. Editing
|
|
113
|
+
requires native write permission and an agent binding that permits edits.
|
|
114
|
+
Prefer `read --start_line N --end_line M` and `patch-page` with `expected_head`:
|
|
115
|
+
pass unique `old_text`/`new_text` excerpts in `edits`, or a contextual single-file
|
|
116
|
+
unified diff in `patch`. Use `--args-file` for JSON arguments. Never replace a
|
|
117
|
+
whole page with the fragment returned by a range read.
|
|
118
|
+
If a read is truncated, continue with `--start_line <next_line>` and
|
|
119
|
+
`--start_column <next_column>`; columns count Unicode characters, one-based.
|
|
120
|
+
This also reads a single line longer than the response limit without gaps.
|
|
121
|
+
In an `apply` batch, `expected_head` is the default read version; each operation
|
|
122
|
+
can override it with its own `expected_head`. All target pages are checked.
|
|
123
|
+
|
|
124
|
+
Keep one `request_id` for an identical retry. Save the returned `task.task_id`;
|
|
125
|
+
on a CLI error the same recovery handle is `data.task_id` with `data.brain`.
|
|
126
|
+
use `edit-status`, `edit-diff`, and `resume-edit` with that ID. Do not split a
|
|
127
|
+
failed merge into unrelated partial edits. One multi-step request should use
|
|
128
|
+
`edit --instruction` or one `apply` batch. Report running, blocked on Review,
|
|
129
|
+
reader sync pending, draft, and completed distinctly. Review is for knowledge
|
|
130
|
+
questions; the task's diff is for inspecting what changed. Resume saved intent
|
|
131
|
+
after its known `review_issue_ids` are resolved. Never claim success before
|
|
132
|
+
checking the task's terminal state and resulting changes.
|
|
133
|
+
|
|
134
|
+
Put related retitle/delete/merge steps in one `apply` batch. Separate
|
|
135
|
+
structural jobs on the same Brain collide with `brain_busy`. Poll the saved
|
|
136
|
+
task and inspect the result; do not ask the user to repeat an authorized step
|
|
137
|
+
just to wait, or promise a timed retry without an active execution. Check the
|
|
138
|
+
actual status of an unrelated busy operation before retrying.
|
|
139
|
+
|
|
140
|
+
## Creating a Brain
|
|
141
|
+
|
|
142
|
+
When the person asks for a new Team Brain, read `gsk team-brain describe --topic create`
|
|
143
|
+
first: it covers the steps (sources, plan, create, build-status, retry, publish), the
|
|
144
|
+
spec passed with `--args-file`, and the two points where the person confirms. The Brain
|
|
145
|
+
belongs to the caller; a GenTeam agent creates one only as a personal agent, on a
|
|
146
|
+
turn its creator started.
|
|
107
147
|
|
|
108
148
|
## Flags
|
|
109
149
|
|
|
@@ -114,27 +154,54 @@ does not grant native editing permission.
|
|
|
114
154
|
| `--scope` | No | For list_available_brains: your Team Brains only. Default 'org'; 'all' is a compatibility alias. (string, one of: org, all) |
|
|
115
155
|
| `--timeout_seconds` | No | For ask and ask-public: server-side deadline, clamped to [5, 570], default 240. Keep the client --timeout above this. (integer) |
|
|
116
156
|
| `--mode` | No | For ask and ask-public: retrieval (default) returns original evidence, facts, caveats and citations without composing a final answer. answer returns a composed response. (string, one of: retrieval, answer, default: `retrieval`) |
|
|
117
|
-
| `--retrieval_strategy` | No | For ask and ask-public in retrieval mode:
|
|
157
|
+
| `--retrieval_strategy` | No | For ask and ask-public in retrieval mode: fast (default) ranks widely without an inner LLM and returns validated original windows for the top files, plus pointers and directory entries to read next. deep progressively searches and selects evidence with an agent; it is slower. The caller checks relevance and expands missing context; neither strategy certifies corpus-wide coverage. (string, one of: fast, deep, default: `fast`) |
|
|
158
|
+
| `--queries` | No | For ask and ask-public with fast retrieval: up to 4 extra search phrasings (each up to 300 characters) searched beside the question, such as the question in the Brain's other language and the specific entity, product or company names involved. (array) |
|
|
159
|
+
| `--pattern` | No | For grep: a regular expression matched case-insensitively against each line of the Brain's published pages, such as 'NTT\|KDDI' or '2026-08\|August 2026'. A pattern starting with its own inline flags, such as (?s), keeps them instead. (string) |
|
|
118
160
|
| `-t`, `--topic` | No | For describe: topic name (ask, public, selection, limits, edit); empty returns the index. (string) |
|
|
119
|
-
| `--page` | No | For read-source, read, edit-page, add-page and
|
|
161
|
+
| `--page` | No | For read-source, read, edit, edit-page, patch-page, add-page and structural edits: the page path as info, tree or a citation lists it, such as people/jane-doe.md. For history: only the versions that changed this page. (string) |
|
|
162
|
+
| `--start_line` | No | For read: first body line, one-based. (integer) |
|
|
163
|
+
| `--start_column` | No | For read: first Unicode character in start_line, one-based. Continue with next_line/next_column. (integer) |
|
|
164
|
+
| `--end_line` | No | For read: last body line, inclusive. (integer) |
|
|
165
|
+
| `--request_id` | No | Stable ID for one edit request. Reuse identical arguments on retry; use resume-edit for a saved task. (string) |
|
|
166
|
+
| `--task_id` | No | Task ID from a prior edit, for edit-status, edit-diff or resume-edit. (string) |
|
|
167
|
+
| `--instruction` | No | For edit: the complete multi-step intent. One bounded AI plan; returns a tracked task and produces a draft. (string) |
|
|
168
|
+
| `--review_issue_ids` | No | Known open knowledge Review issues that must be resolved before this task proceeds. (array) |
|
|
169
|
+
| `--edits` | No | For patch-page or merge-pages (survivor): unique exact excerpts and replacements, all addressing the original body. (array) |
|
|
170
|
+
| `--patch` | No | For patch-page or merge-pages (survivor): single-file unified diff with context; mutually exclusive with edits. Headers target a/<page> and b/<page>. (string) |
|
|
120
171
|
| `--content` | No | For edit-page and add-page: the complete new page body in Markdown, without frontmatter. An entity page keeps its first '# Title' line. For merge-pages: optionally the merged body of the surviving page. (string) |
|
|
121
172
|
| `--title` | No | For retitle-page: the page's new title. (string) |
|
|
122
173
|
| `--new_page` | No | For move-page: the new file name, such as jane-smith (same folder, lowercase slug). (string) |
|
|
123
174
|
| `--into` | No | For merge-pages: the page that absorbs page and survives. (string) |
|
|
124
|
-
| `--expected_head` | No | For edit-page (and apply's default): the head_sha read returned with the page. (string) |
|
|
175
|
+
| `--expected_head` | No | For edit-page, patch-page (and apply's default): the head_sha read returned with the page. (string) |
|
|
125
176
|
| `--ops` | No | For apply: the operations, committed together or not at all. (array) |
|
|
126
177
|
| `--issue_id` | No | For resolve-review: the review item id, such as review-0123456789abcdef. (string) |
|
|
127
178
|
| `--note` | No | For resolve-review: how the conflict was settled, kept with the resolution. (string) |
|
|
128
|
-
| `--label` | No | For rename-brain: the new Brain name. (string) |
|
|
129
|
-
| `--
|
|
130
|
-
| `--
|
|
179
|
+
| `--label` | No | For plan and create: the new Brain's name. For rename-brain: the new Brain name. (string) |
|
|
180
|
+
| `--description` | No | For plan and create: what the Brain is for and who asks it, in a sentence or two; it guides what the build keeps. (string) |
|
|
181
|
+
| `--sources` | No | For plan and create: what the Brain reads, each {"type": ..., fields} as 'describe --topic sources' lists them. Pass label, description and sources with --args-file. (array) |
|
|
182
|
+
| `--spec_hash` | No | For create: the spec_hash plan returned for exactly this spec. (string) |
|
|
183
|
+
| `--build_id` | No | For build-status, retry, publish, set-refresh, access, share and unshare: the build_id create returned. (string) |
|
|
184
|
+
| `--visibility` | No | For publish: personal (only the owner) or team (everyone in the organization who uses Team Brain). (string, one of: personal, team) |
|
|
185
|
+
| `--connector` | No | For sources: whose candidates to list. (string, one of: genteam, microsoft_teams, meetings, box, dropbox, jira, confluence) |
|
|
186
|
+
| `--parent` | No | For sources with box or dropbox: list inside this folder, the parent of an item's list_subfolders; leave out for the top level. (string) |
|
|
187
|
+
| `--cursor` | No | For sources: the next_cursor or next_teams_cursor of the previous page. (string) |
|
|
188
|
+
| `--enabled` | No | For set-refresh: true to refresh on a schedule, false to stop. (boolean) |
|
|
189
|
+
| `--interval_hours` | No | For set-refresh: hours between refreshes; each refresh is a new compile. (integer, one of: 6, 12, 24, 72, 168) |
|
|
190
|
+
| `--agent_id` | No | For bind and unbind: the agent to give the Brain to or take it from (one the person manages). (string) |
|
|
191
|
+
| `--read_only` | No | For bind: true lets the agent only read the Brain; by default it may also edit it. (boolean) |
|
|
192
|
+
| `--email` | No | For share and unshare: the person's address. (string) |
|
|
193
|
+
| `--channel` | No | For share and unshare, instead of --email: a GenTeam channel as access lists it, channel:<server_id>:<channel_id>. (string) |
|
|
194
|
+
| `--role` | No | For share: viewer (reads the Brain) or editor (also edits it; organization members only). (string, one of: viewer, editor) |
|
|
195
|
+
| `--team_id` | No | For sources with microsoft_teams: list this team's channels. (string) |
|
|
196
|
+
| `--folder` | No | For info/tree: list only this folder. For grep: search only under this folder or one page, such as decisions/ or deals/index.md. (string) |
|
|
197
|
+
| `--limit` | No | For info: files per page (default 100, max 500). For read-source: characters per window (default and max 200000; with --pages, per page, default 20000 and at most 200000 in total). For grep: most files (default 20, max 50). For tree: most pages (default 500, max 2000). For history: versions (default 20, max 50). (integer) |
|
|
131
198
|
| `--offset` | No | For info/history/read-source: pass next_offset to continue. (integer) |
|
|
132
199
|
| `--repo` | No | For history: workspace (default, every change) or reader (what people asking the Brain received). (string, one of: workspace, reader) |
|
|
133
200
|
| `--version` | No | For info: expected head_sha when continuing a page. For read-source: a citation's head_sha, to read the page as it saw it; omit for the live page. For read: an earlier page version. For revert: the version to restore. (string) |
|
|
134
|
-
| `--pages` | No | For revert: restore only these pages; omit to roll back every edit since the version. (array) |
|
|
201
|
+
| `--pages` | No | For read-source: up to 8 pages read together at one version, instead of --page. For revert: restore only these pages; omit to roll back every edit since the version. (array) |
|
|
135
202
|
| `--query` | No | For search: literal text to find, case-insensitive. (string) |
|
|
136
203
|
| `--message` | No | For writes: an optional commit message. (string) |
|
|
137
|
-
| `<action>` (positional) | Yes | Action to perform. (string, one of: info, ask, ask-public, read-source, list_available_brains, describe, tree, read, search, review, history, edit-page, add-page, apply, resolve-review, revert, rename-brain, retitle-page, move-page, merge-pages, delete-page) |
|
|
204
|
+
| `<action>` (positional) | Yes | Action to perform. (string, one of: info, ask, ask-public, read-source, grep, list_available_brains, describe, sources, plan, create, build-status, retry, publish, set-refresh, bind, unbind, access, share, unshare, tree, read, search, review, history, edit-page, patch-page, edit, edit-status, resume-edit, edit-diff, add-page, apply, resolve-review, revert, rename-brain, retitle-page, move-page, merge-pages, delete-page) |
|
|
138
205
|
|
|
139
206
|
## See Also
|
|
140
207
|
|
|
@@ -33,7 +33,7 @@ gsk video [options]
|
|
|
33
33
|
| `<query>` (positional) | No | Detailed, self-contained description of the video to generate, written in English. This text is sent to the video model verbatim as the generation prompt, so put all the visual detail here — but NEVER output settings: text like '16:9' or '4 seconds' in the prompt may be rendered literally into the scene. Pass aspect ratio via the `aspect_ratio` parameter and length via `duration`. duration is recommended to be 5 ~ 10 seconds. For image-to-image transitions (user provides start and end frames) e.g., 'A timelapse of a flower blooming in a garden'. Omit when passing the prompt by `query_file` instead. Or omit on a draft_task_id finalize (it reuses the draft's prompt). (string) |
|
|
34
34
|
| `--query_file` | No | Optional. Repo-relative path to a UTF-8 text file containing the prompt, used verbatim in place of `query` (lets you pass a long prompt by path instead of inlining it). Requires `repo_id`. When set, `query` may be omitted. (string) |
|
|
35
35
|
| `--file_name` | No | The name of the video to generate. (string, null) |
|
|
36
|
-
| `-m`, `--model` | No | The model to use for video generation (required for a render; omit only with task_id). kling/v3: Latest Kling V3. Supports native audio, but generates a SILENT video unless audio_enable=true. image_urls are frames, not references: 1 image = first frame (i2v); 2 images = first + last frame (the 2nd is the END frame). No reference mode (use kling/o3 with reference_mode for reference-driven generation). Pro/Standard/Turbo tiers (set via the tier parameter; turbo is the fastest arm but has no audio and no first/last-frame control). Supported ratios: 16:9, 9:16, 1:1. Duration: 3-15s. kling/o3: Kling O3. REQUIRES at least one input image or a video_url (no text-to-video). image_urls are frames (1 = first, 2 = first + last) unless reference_mode=true: style/character references, max 4 total including elements. video_url switches to video-to-video (video_mode 'reference' = motion guide, 'edit' = restyle in footage). Native audio via audio_enable. Pro/Standard tiers. Supported ratios: 16:9, 9:16, 1:1. Duration: 3-15s. gemini/veo3.1: Gemini Veo 3.1 - Latest version with enhanced quality and features. Supports text-to-video and image-to-video generation with improved quality. image_urls are frames unless reference_mode: 1 image = first frame, exactly 2 = first/last frame transition, 3 (or reference_mode=true) = style/character references (max 3, 8s only). video_url (MP4 source) = video EXTENSION mode: continues from the last frame, 7s, up to 1080p. Duration: 4s/6s/8s. Supported ratios: 16:9, 9:16. Supports tier (standard \| fast). Resolution: 720p, 1080p, 4k. minimax/h3: MiniMax H3 2K video model. Supports text-to-video, image-to-video with first/last frame control, and reference-to-video (up to 9 reference images, 3 reference videos, 3 reference audio clips — refer to them in the prompt as 'Image 1', 'Video 1', 'Audio 1', not @-style). Keyframes and references are mutually exclusive. Supported ratios: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-15s. Resolution: 768p or 2K (default). minimax/h3-max: MiniMax H3 Max, a distinct post-trained model with faster generation, stronger prompt adherence, and improved aesthetics. Supports text, first/middle/last frame keyframes, and image/video/audio references — keyframes and references COMBINE on this model (set keyframes via first_frame_url/last_frame_url/middle_image_url + middle_frame_time; the middle frame needs an explicit first_frame_url AND last_frame_url — image_urls are all references here — native 480p/768p, and at least one reference). tier=turbo selects the faster, half-price text/image arm; reference and middle-frame inputs use standard. Duration: 5-15s. Resolution: 480p or 768p. Reference requests also incur a reference-token fee on top of the output charge — a reference video especially can cost more than the output itself. minimax/h3-highspeed: MiniMax H3 Highspeed — H3 inputs and ratios, 480p/768p, 4-15s, faster and cheaper than minimax/h3. minimax/h3-lite: MiniMax H3 Lite — cheapest H3, same inputs, 480p/768p, 4-15s, no adaptive ratio (omit = 16:9). wan/v2.7: Wan v2.7 with enhanced motion smoothness and scene fidelity. Supports text-to-video, image-to-video, reference-to-video, and edit-video. Ratios: 16:9, 9:16, 1:1. Duration: 5s. Resolution: 480p, 720p vidu/q4: Vidu Q4, Vidu's next-generation flagship with native audio. REQUIRES an image (no text-to-video): 1 image_urls entry = first frame (image-to-video, framing follows the image); reference_mode (or 2+ images, or audio_urls) = reference-to-video with up to 12 reference images and 3 voice clips, addressed in the prompt as [@reference_image_1].. / [reference_audio_1].. No end frame, no tiers. Supported ratios (reference mode): 16:9, 9:16, 4:3, 3:4, 1:1. Duration: 3-16s. Resolution: 540p, 720p, 1080p, 2K, 4Krunway/gen4_turbo: A model for generate video with high quality, fast.Supported ratios: 5:3, 3:5. Only support i2v. Duration: 5s, 10spixverse/v6: PixVerse V6 latest model with lifelike motion, richer skin detail, real emotions. Full cinematic control including choreography and camera. Supports text-to-video, image-to-video, transition, and video extend generation. VFX, time-lapse, transformation scenes, product demos, 360° views, multi-shot storytelling. Extend: pass video_url to extend an existing video. Supported ratios: 16:9, 9:16, 4:3, 1:1, 3:4. Duration: 5s, 8s. pixverse/c1: PixVerse C1 with native audio (audio_enable=true), text-to-video, image-to-video, transitions (2 images = first/last frame) and reference-to-video (reference_mode with 1-7 image_urls; address them in the prompt as @image1..@image7). No video extend. Supported ratios: 16:9, 9:16, 4:3, 3:4, 1:1, 3:2, 2:3, 21:9. Duration: 1-15s. Resolution: 360p, 540p, 720p, 1080p. fal-ai/bytedance/seedance-2.0: Bytedance Seedance 2.0 model for highest quality video generation with native audio and lip-sync. Supports text-to-video and image-to-video with first/last frame control. Supported ratios: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 4-15s. Resolution: 480p, 720p, or 1080p on the standard tier (~2.25x cost; fast/mini cap at 720p). tier picks the variant: standard, fast, or mini (cheapest, ~half the standard price). Also supports reference-to-video with up to 9 images, 3 videos, or 3 audio refs via settings.fal-ai/topaz/upscale/video: Topaz Labs Video Upscaler, the preferred upscaler: 2x enhancement of an existing video (source up to 60s) with real detail recovery. Requires video_url parameter. Does NOT generate new videos. Use when user wants to improve quality of existing videos.fal-ai/bytedance-upscaler/upscale/video: ByteDance Video Upscaler for enhancing video quality. Upscales videos to higher resolutions (2k) Requires video_url parameter. Does NOT generate new videos. Use when user wants to improve quality of existing videos.xai/grok-imagine-video: xAI Grok Imagine Video v1.5 for high-quality video generation. Supports text-to-video and image-to-video generation. 2+ image_urls (or reference_mode=true) = style/character references, up to 7 (reference mode caps at 720p). video_url (MP4, 2-15s source) = video EXTENSION mode: continues from the last frame, 2-10s extension, 720p. 480p/720p/1080p output. Supported ratios: 16:9, 4:3, 3:2, 1:1, 2:3, 3:4, 9:16. Duration: 1-15s. tier=lite runs the lightweight low-cost 1.5 Lite arm (text-to-video and image-to-video only — no references, no extension). alibaba/happy-horse: Alibaba Happy Horse video model. High-quality generation from text, or image-to-video from a single first-frame image (one image_url without reference_mode). Supported ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, 9:21, 5:4, 4:5. Duration: 3-15s. Resolution: 720p, 1080p.alibaba/happy-horse/reference-to-video: Alibaba Happy Horse reference-to-video. Generate video from 1-9 reference images; address subjects in the prompt as character1..character9 (order matches image_urls). Requires image_urls. Supported ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, 9:21, 5:4, 4:5. Duration: 3-15s. Resolution: 720p, 1080p.alibaba/happy-horse/video-edit: Alibaba Happy Horse video-edit. Edit a source video using a text prompt; optional 1-5 reference images addressed as @Image1..@Image5. Requires video_url (MP4/MOV, 3-60s). Output preserves source aspect ratio, capped at 15s. Resolution: 720p, 1080p.gemini/omni-flash: Google's multimodal video model with native audio. Supports text-to-video, image-to-video (1 image as first frame), reference-to-video (set reference_mode with image_urls to use images as subject/style references — even a single image; 2+ images are references by default), and video editing (pass video_url to edit an existing clip — restyle, restage, add objects). Bind image roles in the prompt with <FIRST_FRAME> / <IMAGE_REF_0>. Ratios: 16:9, 9:16. Duration: 3-10s. Resolution: 720p. kling/v4: Kling 4.0 — fast 720p/1080p generation. Text-to-video, image-to-video (1 image = start frame, 2 images = start + end frame), reference-to-video (reference_mode or 3+ images: up to 10 image_urls as @Image1.., up to 5 video_urls as @Video1.., up to 7 elements as @Element1..) and video-to-video editing (pass video_url — the prompt is the edit instruction, @Video1 addresses the source clip). No extension mode. Supported ratios: 16:9, 9:16, 1:1, 21:9. Duration: 3-30s. Resolution: 720p, 1080p. tier=flash runs Kling 4.0 Flash (cheaper, faster): same modes, 3-20s at 720p only. bfl/flux-3-preview-high: BFL FLUX 3 video model with native audio. Supports text-to-video and 1-10 ordered keyframes, or continuation from one source video selected with input_mode=start_video. Ratios: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-20s. Resolution: 720p, 1080p, 2K, 4K (2K/4K not available when continuing a video). Only one input family may be used per request.bfl/flux-3-max: FLUX 3 Max, fal's post-trained FLUX 3 video model: top-ranked image-to-video quality, renders a 10s clip in about 12s, native audio. Text-to-video and 1-10 ordered keyframes (2 images = first and last frame). No video continuation, no draft, no auto duration. Ratios: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-20s (explicit). Resolution: 720p, 1080p.fal-ai/bytedance/seedance-2.5: Bytedance Seedance 2.5 for native 30-second single-shot video with synchronized audio. Supports text-to-video and image-to-video with first/last frame control. No tier variants. Supported ratios: auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 4-30s, or 0 = auto (needs video refs). Resolution: 480p, 720p, or 1080p (~2.46x cost). Also supports reference-to-video (up to 30 images, 10 videos, 10 audio refs; no image required), video EDITING and EXTENSION — source in video_urls, the prompt describes the change or continuation.wan/v3.0: Alibaba Wan 3.0 — next-gen open-weight Wan with native audio and single-pass clips up to 30s. Supports text-to-video, image-to-video with first/last frame control, and reference-to-video (reference_mode with up to 10 reference images, 5 videos and 5 audio clips, each ref modality <=15s combined — address them positionally in the prompt as 'Image 1', 'Video 1', 'Audio 1'). tier=fast runs the accelerated Prime arm (~1.4x cost). Supported ratios: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 2-30s. Resolution: 480p, 720p, 1080p. Duration ranges above are hard caps; out-of-range values are clamped to the nearest supported value. (string, one of: kling/v3, kling/o3, gemini/veo3.1, minimax/h3, minimax/h3-max, minimax/h3-highspeed, minimax/h3-lite, wan/v2.7, vidu/q4, runway/gen4_turbo, pixverse/v6, pixverse/c1, fal-ai/bytedance/seedance-2.0, fal-ai/topaz/upscale/video, fal-ai/bytedance-upscaler/upscale/video, xai/grok-imagine-video, alibaba/happy-horse, alibaba/happy-horse/reference-to-video, alibaba/happy-horse/video-edit, gemini/omni-flash, kling/v4, bfl/flux-3-preview-high, bfl/flux-3-max, fal-ai/bytedance/seedance-2.5, wan/v3.0) |
|
|
36
|
+
| `-m`, `--model` | No | The model to use for video generation (required for a render; omit only with task_id). kling/v3: Latest Kling V3. Supports native audio, but generates a SILENT video unless audio_enable=true. image_urls are frames, not references: 1 image = first frame (i2v); 2 images = first + last frame (the 2nd is the END frame). No reference mode (use kling/o3 with reference_mode for reference-driven generation). Pro/Standard/Turbo tiers (set via the tier parameter; turbo is the fastest arm but has no audio and no first/last-frame control). Supported ratios: 16:9, 9:16, 1:1. Duration: 3-15s. kling/o3: Kling O3. REQUIRES at least one input image or a video_url (no text-to-video). image_urls are frames (1 = first, 2 = first + last) unless reference_mode=true: style/character references, max 4 total including elements. video_url switches to video-to-video (video_mode 'reference' = motion guide, 'edit' = restyle in footage). Native audio via audio_enable. Pro/Standard tiers. Supported ratios: 16:9, 9:16, 1:1. Duration: 3-15s. gemini/veo3.1: Gemini Veo 3.1 - Latest version with enhanced quality and features. Supports text-to-video and image-to-video generation with improved quality. image_urls are frames unless reference_mode: 1 image = first frame, exactly 2 = first/last frame transition, 3 (or reference_mode=true) = style/character references (max 3, 8s only). video_url (MP4 source) = video EXTENSION mode: continues from the last frame, 7s, up to 1080p. Duration: 4s/6s/8s. Supported ratios: 16:9, 9:16. Supports tier (standard \| fast). Resolution: 720p, 1080p, 4k. minimax/h3: MiniMax H3 2K video model. Supports text-to-video, image-to-video with first/last frame control, and reference-to-video (up to 9 reference images, 3 reference videos, 3 reference audio clips — refer to them in the prompt as 'Image 1', 'Video 1', 'Audio 1', not @-style). Keyframes and references are mutually exclusive. Supported ratios: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-15s. Resolution: 768p or 2K (default). minimax/h3-max: MiniMax H3 Max, a distinct post-trained model with faster generation, stronger prompt adherence, and improved aesthetics. Supports text, first/middle/last frame keyframes, and image/video/audio references — keyframes and references COMBINE on this model (set keyframes via first_frame_url/last_frame_url/middle_image_url + middle_frame_time; the middle frame needs an explicit first_frame_url AND last_frame_url — image_urls are all references here — native 480p/768p, and at least one reference). tier=turbo selects the faster, half-price text/image arm; reference and middle-frame inputs use standard. Duration: 5-15s. Resolution: 480p or 768p. Reference requests also incur a reference-token fee on top of the output charge — a reference video especially can cost more than the output itself. minimax/h3-highspeed: MiniMax H3 Highspeed — H3 inputs and ratios, 480p/768p, 4-15s, faster and cheaper than minimax/h3. minimax/h3-lite: MiniMax H3 Lite — cheapest H3, same inputs, 480p/768p, 4-15s, no adaptive ratio (omit = 16:9). wan/v2.7: Wan v2.7 with enhanced motion smoothness and scene fidelity. Supports text-to-video, image-to-video, reference-to-video, and edit-video. Ratios: 16:9, 9:16, 1:1. Duration: 5s. Resolution: 480p, 720p vidu/q4: Vidu Q4, Vidu's next-generation flagship with native audio. REQUIRES an image (no text-to-video): 1 image_urls entry = first frame (image-to-video, framing follows the image); reference_mode (or 2+ images, or audio_urls) = reference-to-video with up to 12 reference images and 3 voice clips, addressed in the prompt as [@reference_image_1].. / [reference_audio_1].. No end frame, no tiers. Supported ratios (reference mode): 16:9, 9:16, 4:3, 3:4, 1:1. Duration: 3-16s. Resolution: 540p, 720p, 1080p, 2K, 4Krunway/gen4_turbo: A model for generate video with high quality, fast.Supported ratios: 5:3, 3:5. Only support i2v. Duration: 5s, 10spixverse/v6: PixVerse V6 latest model with lifelike motion, richer skin detail, real emotions. Full cinematic control including choreography and camera. Supports text-to-video, image-to-video, transition, and video extend generation. VFX, time-lapse, transformation scenes, product demos, 360° views, multi-shot storytelling. Extend: pass video_url to extend an existing video. Supported ratios: 16:9, 9:16, 4:3, 1:1, 3:4. Duration: 5s, 8s. pixverse/c1: PixVerse C1 with native audio (audio_enable=true), text-to-video, image-to-video, transitions (2 images = first/last frame) and reference-to-video (reference_mode with 1-7 image_urls; address them in the prompt as @image1..@image7). No video extend. Supported ratios: 16:9, 9:16, 4:3, 3:4, 1:1, 3:2, 2:3, 21:9. Duration: 1-15s. Resolution: 360p, 540p, 720p, 1080p. fal-ai/bytedance/seedance-2.0: Bytedance Seedance 2.0 model for highest quality video generation with native audio and lip-sync. Supports text-to-video and image-to-video with first/last frame control. Supported ratios: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 4-15s. Resolution: 480p, 720p, or 1080p on the standard tier (~2.25x cost; fast/mini cap at 720p). tier picks the variant: standard, fast, or mini (cheapest, ~half the standard price). Also supports reference-to-video with up to 9 images, 3 videos, or 3 audio refs via settings.fal-ai/topaz/upscale/video: Topaz Labs Video Upscaler, the preferred upscaler: 2x enhancement of an existing video (source up to 60s) with real detail recovery. Requires video_url parameter. Does NOT generate new videos. Use when user wants to improve quality of existing videos.fal-ai/bytedance-upscaler/upscale/video: ByteDance Video Upscaler for enhancing video quality. Upscales videos to higher resolutions (2k) Requires video_url parameter. Does NOT generate new videos. Use when user wants to improve quality of existing videos.xai/grok-imagine-video: xAI Grok Imagine Video v1.5 for high-quality video generation. Supports text-to-video and image-to-video generation. 2+ image_urls (or reference_mode=true) = style/character references, up to 7 (reference mode caps at 720p). video_url (MP4, 2-15s source) = video EXTENSION mode: continues from the last frame, 2-10s extension, 720p. 480p/720p/1080p output. Supported ratios: 16:9, 4:3, 3:2, 1:1, 2:3, 3:4, 9:16. Duration: 1-15s. tier=lite runs the lightweight low-cost 1.5 Lite arm (text-to-video and image-to-video only — no references, no extension). alibaba/happy-horse: Alibaba Happy Horse video model. High-quality generation from text, or image-to-video from a single first-frame image (one image_url without reference_mode). Supported ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, 9:21, 5:4, 4:5. Duration: 3-15s. Resolution: 720p, 1080p.alibaba/happy-horse/reference-to-video: Alibaba Happy Horse reference-to-video. Generate video from 1-9 reference images; address subjects in the prompt as character1..character9 (order matches image_urls). Requires image_urls. Supported ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, 9:21, 5:4, 4:5. Duration: 3-15s. Resolution: 720p, 1080p.alibaba/happy-horse/video-edit: Alibaba Happy Horse video-edit. Edit a source video using a text prompt; optional 1-5 reference images addressed as @Image1..@Image5. Requires video_url (MP4/MOV, 3-60s). Output preserves source aspect ratio, capped at 15s. Resolution: 720p, 1080p.gemini/omni-1.2-flash: Gemini Omni 1.2 Flash — Google's multimodal video model with native audio: 3-15s, 16:9/9:16, 360p/720p/1080p/4k; 1 image = first frame, 2 = first + last frame (3-10s), reference_mode = up to 6 reference images + 3 clips (<=3s, describe roles in plain words); video_url = edit that clip (video_mode=extend continues it by 3-10s).kling/v4: Kling 4.0 — fast 720p/1080p generation. Text-to-video, image-to-video (1 image = start frame, 2 images = start + end frame), reference-to-video (reference_mode or 3+ images: up to 10 image_urls as @Image1.., up to 5 video_urls as @Video1.., up to 7 elements as @Element1..) and video-to-video editing (pass video_url — the prompt is the edit instruction, @Video1 addresses the source clip). No extension mode. Supported ratios: 16:9, 9:16, 1:1, 21:9. Duration: 3-30s. Resolution: 720p, 1080p. tier=flash runs Kling 4.0 Flash (cheaper, faster): same modes, 3-20s at 720p only. bfl/flux-3-preview-high: BFL FLUX 3 video model with native audio. Supports text-to-video and 1-10 ordered keyframes, or continuation from one source video selected with input_mode=start_video. Ratios: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-20s. Resolution: 720p, 1080p, 2K, 4K (2K/4K not available when continuing a video). Only one input family may be used per request.bfl/flux-3-max: FLUX 3 Max, fal's post-trained FLUX 3 video model: top-ranked image-to-video quality, renders a 10s clip in about 12s, native audio. Text-to-video and 1-10 ordered keyframes (2 images = first and last frame). No video continuation, no draft, no auto duration. Ratios: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 5-20s (explicit). Resolution: 720p, 1080p.fal-ai/bytedance/seedance-2.5: Bytedance Seedance 2.5 for native 30-second single-shot video with synchronized audio. Supports text-to-video and image-to-video with first/last frame control. No tier variants. Supported ratios: auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 4-30s, or 0 = auto (needs video refs). Resolution: 480p, 720p, or 1080p (~2.46x cost). Also supports reference-to-video (up to 30 images, 10 videos, 10 audio refs; no image required), video EDITING and EXTENSION — source in video_urls, the prompt describes the change or continuation.wan/v3.0: Alibaba Wan 3.0 — next-gen open-weight Wan with native audio and single-pass clips up to 30s. Supports text-to-video, image-to-video with first/last frame control, and reference-to-video (reference_mode with up to 10 reference images, 5 videos and 5 audio clips, each ref modality <=15s combined — address them positionally in the prompt as 'Image 1', 'Video 1', 'Audio 1'). tier=fast runs the accelerated Prime arm (~1.4x cost). Supported ratios: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16. Duration: 2-30s. Resolution: 480p, 720p, 1080p. Duration ranges above are hard caps; out-of-range values are clamped to the nearest supported value. (string, one of: kling/v3, kling/o3, gemini/veo3.1, minimax/h3, minimax/h3-max, minimax/h3-highspeed, minimax/h3-lite, wan/v2.7, vidu/q4, runway/gen4_turbo, pixverse/v6, pixverse/c1, fal-ai/bytedance/seedance-2.0, fal-ai/topaz/upscale/video, fal-ai/bytedance-upscaler/upscale/video, xai/grok-imagine-video, alibaba/happy-horse, alibaba/happy-horse/reference-to-video, alibaba/happy-horse/video-edit, gemini/omni-1.2-flash, kling/v4, bfl/flux-3-preview-high, bfl/flux-3-max, fal-ai/bytedance/seedance-2.5, wan/v3.0) |
|
|
37
37
|
| `-i`, `--image_urls` | No | The URLs of the images to use as reference key frames for the video generation. For single image models, provide 1 image. For multi-image models (like kling/v1.6/pro/elements), provide 1-4 images in sequence order. For start-end models (like vidu, pixverse), provide exactly 2 images (start and end frames). (default is [], if the task is based on one or more reference images, it is required) IMPORTANT: without reference_mode, the first image becomes the video's literal opening frame — identity markers visible in it (flags, jerseys, text, logos, faces) persist in the output and the text prompt CANNOT override them. When generating a batch of subject-specific clips (per-country / per-person / per-product), each clip needs an image matching its own subject — never reuse another subject's image. (array) |
|
|
38
38
|
| `--first_frame_url` | No | MiniMax H3 Max keyframe control: image URL the video opens on exactly, ALONGSIDE reference_mode references (the output canvas follows this image). Without reference_mode the first image_urls entry already is the first frame; in reference_mode (including any middle-frame request) it is REQUIRED, because image_urls entries are all subject references. H3 Max only: other models ignore it, and base minimax/h3 rejects it (frames there come from image_urls). (string) |
|
|
39
39
|
| `--middle_image_url` | No | MiniMax H3 Max keyframe control: image the video passes through at middle_frame_time. Requires an explicit first frame (first_frame_url) AND last frame (last_frame_url) — a middle frame forces reference mode, where every image_urls entry is a subject reference — plus native 480p/768p and at least one reference image/video/audio. H3 Max only: other models ignore it, and base minimax/h3 rejects it. (string) |
|
|
@@ -43,7 +43,7 @@ gsk video [options]
|
|
|
43
43
|
| `-d`, `--duration` | No | The duration in seconds. 0 = auto (Seedance 2.5 with video_urls; required for edit prompts — output keeps the source length; extensions: pass the new segment's length). (number, default: `5`) |
|
|
44
44
|
| `-a`, `--audio_url` | No | Audio URL for audio integration in video generation. - Optional for Wan v2.5 model (custom audio: music or speech, WAV/MP3, 3-10s, up to 15MB) - Required for OmniHuman model (audio-driven animation) Leave empty if no audio is needed. (string, default: ``) |
|
|
45
45
|
| `--video_url` | No | Source video URL for video extension or video-to-video. On Grok Imagine Video and Gemini Veo 3.1 it switches to EXTENSION mode; on Kling O3 it switches to video-to-video (see video_mode). Supports both web URLs and AI Drive paths (starting with '/' or 'aidrive://') (string, default: ``) |
|
|
46
|
-
| `--video_mode` | No | How video_url is used. Kling O3: 'reference' (default — motion/camera guide) or 'edit' (restyle inside the source footage); no extend. Gemini Veo 3.1 / Grok Imagine Video: 'extend' only. Ignored without video_url. (string, one of: reference, edit, extend) |
|
|
46
|
+
| `--video_mode` | No | How video_url is used. Kling O3: 'reference' (default — motion/camera guide) or 'edit' (restyle inside the source footage); no extend. Gemini Veo 3.1 / Grok Imagine Video: 'extend' only. Gemini Omni 1.2 Flash: 'edit' (default) or 'extend'. Ignored without video_url. (string, one of: reference, edit, extend) |
|
|
47
47
|
| `--keep_audio` | No | Kling O3 video-to-video only: keep the source video's audio track in the output. (boolean) |
|
|
48
48
|
| `--input_mode` | No | FLUX 3 Video only: set start_video to continue video_url from its final frames. Omit for all other models and for text/image input. (string, one of: start_video) |
|
|
49
49
|
| `--video_urls` | No | Reference video URLs for reference-to-video models. For Wan v2.7 and Seedance 2.0 reference-to-video: provide 1-3 reference videos and refer to them in the prompt as @Video1, @Video2, @Video3. Seedance 2.0 also needs reference_mode plus at least one image_url; combined video duration 2-15s. For Seedance 2.5: up to 10 videos (≤30s combined), no image required; also the edit/extension source (edit: duration=0, output keeps the source length; extension: duration = the new segment's length). For MiniMax H3: up to 3 motion reference videos (each 2-15s, ≤15s combined), referred to in the prompt as 'Video 1' etc. (array, default: `[]`) |
|
|
@@ -51,7 +51,7 @@ gsk video [options]
|
|
|
51
51
|
| `--repo_id` | No | Optional. Second Brain repo id (provided in the agent's system prompt). Required only when image_urls / audio_url / video_url / video_urls contain relative paths from a Second Brain project (e.g. 'assets/clip.mp4'). Absolute http(s) URLs and data: URLs do not need this parameter. (string) |
|
|
52
52
|
| `--audio_enable` | No | Set true when the user wants audio, voiceover, or narration. kling/v3 outputs a SILENT video unless true (audio costs more per second). Omit to keep each model's default (kling: off; seedance/veo: on). (boolean) |
|
|
53
53
|
| `--tier` | No | Model tier (options are model-specific — check get_model_info tier_options). Seedance 2.0: standard \| fast (quicker/cheaper, 720p cap) \| mini (cheapest, ~half the standard price, 720p cap). Veo 3 / Veo 3.1 family: standard \| fast (quicker, lower cost). Kling V3: pro \| standard \| turbo (fastest, no audio/frame control). Kling O3/Motion Control: pro \| standard. Kling 4.0: standard \| flash (cheaper/faster, 3-20s, 720p only). Vidu Q3: standard \| turbo (half price). Wan 3.0: standard \| fast (the accelerated Prime arm — quicker but ~1.4x the price). MiniMax H3 Max: standard \| turbo (faster, half price; references use standard). MiniMax H3 ignores tier. Models without tiers ignore it. (string, one of: standard, fast, mini, pro, turbo, max, flash) |
|
|
54
|
-
| `--video_size` | No | Output resolution. 'auto' uses the model default (usually 720p). MiniMax H3 defaults to 2K with a cheaper 768p tier; MiniMax H3 Max uses 480p or 768p. Values a model does not publish fall back to its default. Higher tiers bill more per second — see each model's Resolution line. Default: auto. (string, one of: auto, 480p, 720p, 768p, 1080p, 2K, 4k, default: `auto`) |
|
|
54
|
+
| `--video_size` | No | Output resolution. 'auto' uses the model default (usually 720p). MiniMax H3 defaults to 2K with a cheaper 768p tier; MiniMax H3 Max uses 480p or 768p. Values a model does not publish fall back to its default. Higher tiers bill more per second — see each model's Resolution line. Default: auto. (string, one of: auto, 360p, 480p, 720p, 768p, 1080p, 2K, 4k, default: `auto`) |
|
|
55
55
|
| `--draft` | No | Seedance 2.5 only: render a 480p draft (billed at the 480p rate) to check structure, motion and prompt fit; a delivered draft renders its final 1080p via draft_task_id (same content, no new inputs). Ignored on other models. (boolean, default: `False`) |
|
|
56
56
|
| `--draft_task_id` | No | task_id of a successful Seedance 2.5 draft (within 7 days): renders its final 1080p video, reusing the draft's prompt, references, duration and ratio. Send no other inputs and no draft flag with it; bills the 1080p rate. 1080p is the only final size the draft API accepts (a 720p ask re-renders fresh, no pointer). (string) |
|
|
57
57
|
| `--estimate` | No | Dry run: return the credit estimate for exactly these settings (model, duration, resolution, tier, audio, draft) and generate nothing - nothing is billed. Run it with the same flags before a paid render when you must decide whether to confirm the spend first. (boolean, default: `False`) |
|