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 hedra-character-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/hedra-character-video)<a href="https://agentmods.dev/skills/calesthio/generative-media-skills/hedra-character-video"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/hedra-character-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/hedra-character-video"><img src="https://agentmods.dev/badge/skills/calesthio/generative-media-skills/hedra-character-video.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 16 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
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 | $0.00078 | $0.04462 |
| Opus 5 | $0.00039 | $0.02231 |
| Sonnet 5 | $0.00016 | $0.00892 |
| Haiku 4.5 | $0.00008 | $0.00446 |
Grade A, and why
hedra-character-video scanned grade A with 0 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 268 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hedra character video production
Use Hedra when the job is fundamentally a character performance: a still or generated character needs to speak, sing, react, gesture, host a lesson, sell a product, or participate as a real-time avatar. Treat Hedra as an audio-driven performance system first, not as a generic cinematic video model.
Volatile facts in this skill were verified from Hedra first-party pages on 2026-07-10. Re-check Hedra's model list, pricing, duration limits, subscription/API access, and legal pages before paid or compliance-sensitive production.
Documented facts to anchor decisions
Verified 2026-07-10:
- Hedra's public API base URL in the docs is
https://api.hedra.com/web-app/public, and calls require anX-API-Keyheader. API access documentation says users need a Hedra account, a paid plan for API access, and an API key from the API profile page. - The API avatar guide documents two Hedra-native avatar models:
- Hedra Avatar, model ID
26f0fc66-152b-40ab-abed-76c43df99bc8, for talking-head/lip-sync clips up to 10 minutes. - Hedra Omnia, model ID
ab372b84-432f-44f5-bacc-c2542465f712, for motion avatar videos with full-body movement up to 8 seconds.
- Hedra Avatar, model ID
- Both documented avatar models require an image and audio input. Avatar video length is driven by audio length; long-form avatar generation should poll status for
progressandeta_sec. - The app workflow for UGC avatar videos accepts an image from upload or Hedra generation, audio from upload/recording/AI generation, a chosen avatar model, aspect ratio, resolution, and batch size. Hedra's own app tips favor front-facing portraits, high resolution, clean audio, and short tests before long runs.
- File specs list
.wavand.mp3audio, a 10 MB maximum file size for all file types, a free-plan max audio length of 20 seconds, and paid-plan max audio/video length up to 10 minutes depending on model. The same page lists Hedra Avatar at 10 minutes and Hedra Omnia at 8 seconds. - Hedra's video model guide frames Hedra Avatar as audio-driven avatar work with image + audio input, reliable lip sync, long-form dialogue, and a 7 credits/second listing; it frames Hedra Omnia as short cinematic dialogue with text + audio input, camera control, lifelike motion/emotion, and a 15 credits/second listing. Treat these credit figures as volatile.
- Hedra API generation supports
type: "video",ai_model_id,start_keyframe_idorstart_keyframe_url,audio_idor inlineaudio_generation,generated_video_inputs.text_prompt,aspect_ratio,resolution,duration_ms,batch_size, and status polling at/generations/{generation_id}/status. - Hedra API text-to-speech supports
type: "text_to_speech",voice_id,text, optionalstability,speed, andlanguage. The app docs mention ElevenLabs V3, ElevenLabs V2, and MiniMax as selectable audio models and speech-control tags for tone, reaction, volume/energy, and pacing/rhythm. - Hedra documents voice cloning by uploading an audio sample, submitting a
voice_clonegeneration, and using the returned voice asset asvoice_id. - Hedra Studio documents speaker selection for multi-character images: place and size a speaker indicator over the character that should speak. The selected character speaks, sings, or performs from the audio input.
- Hedra Live Avatar is a separate real-time path using LiveKit plus Hedra, with environment variables for LiveKit, Hedra, and often OpenAI. The guide documents
@hedra/create-hedra-avatar,livekit-agents[hedra]~=1.0,HEDRA_AVATAR_ID, and session behavior in LiveKit rooms. - Hedra's acceptable-use policy prohibits illegal use, harming minors, misleading false content, personal information about people other than yourself except for specific historical/public-right contexts, non-consensual impersonation/deepfakes, sensitive personal information, sexual explicit content, harassment/defamation, and several high-stakes automated decision or discriminatory uses.
- Hedra's privacy and biometric policies say uploaded/generated user content can include photos, audio/voice, videos, metadata, and derived data; avatar/image/audio features may derive biometric data such as facial geometry templates or information from audio/voice recordings; biometric data is used to provide features and for legal compliance, with consent where required by applicable biometric law.
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 · 268 lines · 78 tokens per session scan A 0c015969519b
hedra-character-video is a skill published in the GitHub repository calesthio/generative-media-skills (170 stars, last pushed 2mo ago), licensed MIT. It adds 78 tokens to every session and 4,462 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. 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…