Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add calesthio/generative-media-skills --skill minimax-hailuo-videogit clone --depth 1 https://github.com/calesthio/generative-media-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/calesthio/generative-media-skills/minimax-hailuo-video)<a href="https://agentmods.dev/skills/calesthio/generative-media-skills/minimax-hailuo-video"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/minimax-hailuo-video/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/calesthio/generative-media-skills/minimax-hailuo-video"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/minimax-hailuo-video.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00128 | $0.12394 |
| Opus 5 | $0.00064 | $0.06197 |
| Sonnet 5 | $0.00026 | $0.02479 |
| Haiku 4.5 | $0.00013 | $0.01239 |
Grade C, and why
minimax-hailuo-video scanned grade C with 3 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Instruction-override phrasinghighPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
Use only prompts and media you are authorized to submit. For identifiable people, record informed consent for this purpose, territory, audience, duration, editing, and synthetic transformation; obtain any publicity, perf Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
import argparse, base64, hashlib, http.client, ipaddress, json, os, pathlib, random, socket, ssl, subprocess, sys, tempfile, threading, time, urllib.error, urllib.parse, urllib.request Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
process = subprocess.Popen(argv, stdout=subprocess.PIPE, stderr=subprocess.PIPE) How it starts
The opening of the file, as written. The whole thing — 656 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MiniMax Hailuo video
Build a reviewed request, obtain exact cost approval, create at most one paid task, then preserve the result and its evidence. Treat documentation gaps as unknowns, not defaults.
Start with the boundary
Choose one service region and keep its account, key, billing, host, and terms together:
| Region | Console/docs | API host | Billing currency |
|---|---|---|---|
| Global | platform.minimax.io |
https://api.minimax.io |
USD |
| Mainland China | platform.minimaxi.com |
https://api.minimaxi.com |
CNY |
The two platforms are separate service regions. A key or plan bought on one must not be assumed to work on the other. A 401 can be a wrong-region symptom. Use a pay-as-you-go API key for the examples below; a Token Plan/Subscription Key has separate quotas and model access.
Keep these products out of this workflow:
hailuoai.videois the consumer Hailuo product with separate terms, privacy behavior, features, and commercial-use implications.- Video Agent/template generation is a different template API. Its deprecated query endpoint and nine-hour link statement do not define general video generation.
- Image, speech, music, and text endpoints are separate modalities. General Hailuo video generation does not document an audio input, audio prompt, or generated-audio control.
- Gateways and resellers have their own schemas, retention, pricing, and support. Do not call them “the MiniMax API.â€
Route by mode before choosing a model
Current first-party general video generation uses POST /v1/video_generation for four modes. The API overview highlights the Hailuo 2.3 family and Hailuo-02; detailed endpoint pages also enumerate older models.
| Mode | Required media field(s) | Documented models | Documented output choices |
|---|---|---|---|
| Text-to-video (T2V) | none; prompt required |
MiniMax-Hailuo-2.3, MiniMax-Hailuo-02; T2V endpoint also lists T2V-01-Director, T2V-01 |
2.3/02: 768P at 6 or 10 s; 1080P at 6 s; 24 fps |
| Image-to-video (I2V) | first_frame_image |
MiniMax-Hailuo-2.3, MiniMax-Hailuo-2.3-Fast, MiniMax-Hailuo-02; endpoint also lists I2V-01-Director, I2V-01-live, I2V-01 |
2.3/Fast/02: 768P at 6 or 10 s; 1080P at 6 s. 02 also: 512P at 6 or 10 s; 24 fps |
| First/last frame | first_frame_image, last_frame_image |
MiniMax-Hailuo-02 |
768P at 6 or 10 s; 1080P at 6 s; 24 fps. The endpoint explicitly says 512P is unsupported |
| Subject reference | subject_reference array |
S2V-01 |
Unknown: its endpoint omits duration, resolution, and current price; the general guide example nevertheless sends 6 s and 1080P |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 656 lines · 128 tokens per session scan C 484eebb04aa6
minimax-hailuo-video is a skill published in the GitHub repository calesthio/generative-media-skills (171 stars, last pushed 2mo ago), licensed MIT. It adds 128 tokens to every session and 12,394 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it C with 3 findings (instruction-override phrasing, makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
cliptalk-cover-director
Produces evidence-backed cover candidates and reviewable cover variants for a ClipTalk video. Use when the user asks for a cover, poster frame, thumbnail, or multiple cover directions; do not use for timeline editing or social-video reframing.
cliptalk-smart-reframe
Creates a subject-aware, time-varying crop track and a review-only social-format preview from an accepted ClipTalk cut. Use for automatic vertical, square, or portrait reframing; do not use for a fixed manual crop or before content editing is accepted.
cliptalk-content-extractor
Locates and assembles source passages matching a semantic request. Use for extracting explanations, topics, quotes, demonstrations, or other specifically described content.
cliptalk-interview-editor
Produces a coherent interview edit by combining speaker discovery, topic selection, dialogue context, cleanup, subtitles, and preview. Use for interviews, podcasts, testimonials, or question-and-answer recordings.
cliptalk-shortform-hook-director
Finds and assembles a reviewable short-form cut with a strong opening hook. Use for Shorts, Reels, social clips, talking-head cutdowns, or requests for a punchier opening.
cliptalk-social-reframe-exporter
Creates a review-only 9:16, 4:5, 1:1, or 16:9 version from an existing accepted ClipTalk cut, then checks the rendered preview. Use only when a cut already exists and the user asks to adapt it for Shorts, Reels, Douyin, Xiaohongshu, WeChat Channels, or square feeds; do not use when the user still needs content found…