📦
youtube-thumbnail-design
77Design High-CTR YouTube Thumbnails
YouTube thumbnails must be clear at mobile feed size and still earn clicks. This skill gives practical rules for layout, color, text, faces, and A/B testing.
Design & Creative View
Skills ai-avatar-video
📦
ai-avatar-video
Content revision r1 Medium Risk ⚙️ External commands🌐 Network access
Create AI Avatar Videos
Creating avatar videos often requires model selection, voice setup, and multi-step media workflows. This skill guides Claude, Codex, and Claude Code through inference.sh avatar video commands and safe input planning.
Supports: Claude Codex Code(CC)
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "ai-avatar-video" from https://skillstore.io/skills/halt-catch-fire-ai-avatar-video.md and its manifest at https://skillstore.io/api/skills/halt-catch-fire-ai-avatar-video/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "ai-avatar-video". A marketing manager asks for a product demo from one approved portrait and a 45-second script.
Expected outcome:
The skill recommends P-Video-Avatar, lists required portrait and script inputs, selects a voice, and warns that media is sent to inference.sh.
Using "ai-avatar-video". A course creator needs the same lesson intro in English and Spanish.
Expected outcome:
The skill suggests built-in TTS when possible, or a separate speech step, then a lip sync step for each language.
Using "ai-avatar-video". A social team wants three short UGC-style clips with different presenters.
Expected outcome:
The skill outlines a batch workflow with separate voices, approved images, consistent resolution, and clear labeling as synthetic media.
Security Audit
Medium Riskv4 • 7/5/2026 Open versioned report
No prompt injection attempt was found in SKILL.md. The skill intentionally relies on the belt CLI and external inference services, so confirmed risks center on command execution, third-party media processing, and synthetic presenter misuse.
1
Files scanned
268
Lines analyzed
16
Review items
0
False positives ignored
Confirmed security concerns (3)
Medium
Third-Party Media Processing
The workflows send image, audio, video URLs, and scripts to inference.sh apps through belt. Private media or likeness data can leave the local environment.
Multiple documented workflows submit user media URLs and scripts to remote inference apps. This is core behavior, not a hypothetical path.
Medium
Broad Belt CLI Permission
The frontmatter allows Bash(belt *), which permits any belt subcommand rather than only listed avatar commands. This broad scope increases misuse impact.
The allowed-tools declaration explicitly permits any belt subcommand, and the skill encourages browsing all video apps through belt.
Medium
Synthetic Testimonial Misuse Risk
The UGC workflows create synthetic presenters and testimonial-style scripts. Without consent and disclosure controls, these videos can mislead audiences.
The source explicitly describes UGC marketing content, batch presenters, and testimonial-style scripts. Legitimate use is possible, but misuse risk is clear.
Capability review items (16)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Medium
Ruby/shell backtick execution
> **Install the belt CLI skill:** `npx skills add belt-sh/cli`
This line instructs installing a third-party CLI skill with npx, which is a supply-chain and external-command action. The command is explicit and user-visible.
Medium
Ruby/shell backtick execution
```bash
The bash block instructs users to run belt login and belt app run, which executes an external CLI and sends inputs to remote inference services.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The workflow runs belt commands for text-to-speech and avatar generation, invoking remote apps through an external CLI. This is expected but still security-relevant.
Medium
Ruby/shell backtick execution
```bash
The dubbing workflow runs multiple belt commands against external media-processing apps. It can transmit video, transcript, and audio data outside the local environment.
Medium
Ruby/shell backtick execution
```bash
The bash block runs belt app run for an external inference app with user-provided media or script inputs. This is expected functionality but requires trust in the CLI and provider.
Medium
Ruby/shell backtick execution
```bash
The batch example uses a shell loop and belt commands to automate remote app calls. The loop uses a bounded local voice list, but it is still external command execution.
Medium
Ruby/shell backtick execution
```bash
The related-skills block instructs npx skills add for additional skills, which can install external content from a GitHub source. This is a supply-chain risk.
Low
Hardcoded URL
. [Install instructions](https://raw.githubusercontent.com/infer
The link points to raw GitHub install instructions for an external CLI dependency. Following it can expose users to supply-chain content outside the skill file.
Risk Factors
⚙️ External commands (33)
SKILL.md:7 SKILL.md:17 SKILL.md:19-28 SKILL.md:28-36 SKILL.md:36-37 SKILL.md:37-38 SKILL.md:38-39 SKILL.md:39-55 SKILL.md:55-63 SKILL.md:63-67 SKILL.md:67-76 SKILL.md:76-80 SKILL.md:80-85 SKILL.md:85-91 SKILL.md:91-104 SKILL.md:104-108 SKILL.md:108-113 SKILL.md:113-119 SKILL.md:119-124 SKILL.md:124-128 SKILL.md:128-133 SKILL.md:133-139 SKILL.md:139-152 SKILL.md:152-158 SKILL.md:158-172 SKILL.md:172-178 SKILL.md:178-194 SKILL.md:194-206 SKILL.md:206-218 SKILL.md:218-235 SKILL.md:235 SKILL.md:241-259 SKILL.md:259-261
Audited by: codex View Audit History →
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/halt-catch-fire-ai-avatar-video/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/halt-catch-fire-ai-avatar-video?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/halt-catch-fire-ai-avatar-video?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/halt-catch-fire-ai-avatar-video/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/halt-catch-fire-ai-avatar-video.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
halt-catch-fire. (2026). ai-avatar-video security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/halt-catch-fire-ai-avatar-video/audits/4BibTeX citation
@techreport{halt-catch-fire-halt-catch-fire-ai-avatar-video-2026,
author = {halt-catch-fire},
title = {ai-avatar-video security audit report (audit version 4)},
institution = {Skillstore},
year = {2026},
number = {4},
url = {https://skillstore.io/skills/halt-catch-fire-ai-avatar-video/audits/4},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "ai-avatar-video security audit report (audit version 4)"
version: "unspecified"
type: report
authors:
- name: "halt-catch-fire"
date-released: "2026-07-05"
url: "https://skillstore.io/skills/halt-catch-fire-ai-avatar-video/audits/4"
identifiers:
- type: other
value: "skillstore:halt-catch-fire-ai-avatar-video:audit:4"
description: "Skillstore immutable audit report identifier"
Compare variants
11 installable variantsEach author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.
Why this variant is first
Higher Skillstore usage
Author / source Skillstore Score Evidence Confidence Skillstore usage Updated
runcomfy-com Recommended
runcomfy-com-ai-avatar-video
Skillstore Score 79
Evidence Confidence High
Skillstore usage 9
Updated
2026-08-21
inference-sh-skills-ai-avatar-video
Skillstore Score 79
Evidence Confidence High
Skillstore usage 5
Updated
2026-08-21
infsh-skills-ai-avatar-video
Skillstore Score 69
Evidence Confidence High
Skillstore usage 12
Updated
2026-08-21
inference-skills-ai-avatar-video
Skillstore Score 69
Evidence Confidence High
Skillstore usage 10
Updated
2026-08-21
skills-shell-ai-avatar-video
Skillstore Score 50
Evidence Confidence High
Skillstore usage 24
Updated
2026-08-21
halt-catch-fire Current
halt-catch-fire-ai-avatar-video
Skillstore Score 50
Evidence Confidence High
Skillstore usage 13
Updated
2026-08-21
qu-skills-ai-avatar-video
Skillstore Score 50
Evidence Confidence Medium
Skillstore usage 9
Updated
2026-08-21
agentspace-so-ai-avatar-video
Skillstore Score 38
Evidence Confidence High
Skillstore usage 7
Updated
2026-08-21
doany-ai-ai-avatar-video
Skillstore Score 38
Evidence Confidence Medium
Skillstore usage 10
Updated
2026-08-21
inference-sh-ai-avatar-video
Skillstore Score 38
Evidence Confidence Medium
Skillstore usage 9
Updated
2026-08-21
101-skills-ai-avatar-video
Skillstore Score 38
Evidence Confidence Medium
Skillstore usage 6
Updated
2026-08-21
Skillstore Score
Why this score Evidence Confidence: High 55
Architecture
85
Maintainability
87
Content
67
Community
91
Spec Compliance
What You Can Build
Create product presenter videos
Marketing teams can turn approved portraits and scripts into short product videos with virtual presenters.
Produce lesson introductions
Educators can create talking head introductions for course modules using selected voices and languages.
Localize existing videos
Localization teams can transcribe, translate, generate speech, and lip sync videos for new audiences.
Try These Prompts
Create a basic avatar clip
Create a 30-second avatar video from this portrait URL and script. Use P-Video-Avatar with built-in TTS and a friendly English voice.
Choose the best model
Compare P-Video-Avatar, OmniHuman, Fabric, and PixVerse for a one-minute product explainer. Recommend the best model and explain cost, speed, and audio needs.
Plan a full production workflow
Plan a workflow to generate a portrait, create expressive speech, and produce a vertical avatar video for a product launch. Include inputs I must prepare.
Design a batch campaign
Design a batch process for three presenters and two languages using approved portraits and scripts. Include consent checks and where each output is used.
Best Practices
- Use only portraits and voices where you have consent and usage rights.
- Review generated videos for accuracy, disclosure, and brand safety before publishing.
- Use clear media URLs and keep private assets out of public links when possible.
Avoid
- Do not create videos that impersonate real people without authorization.
- Do not send confidential customer media to external services without approval.
- Do not install related skills or CLIs from unreviewed sources during a production task.
Frequently Asked Questions
Does this skill create videos locally?
No. It uses the inference.sh belt CLI and external models, so network access is required.
Which model does it recommend first?
It recommends P-Video-Avatar because the source says it is faster, cheaper, and includes text to speech.
Can it use my own audio?
Yes. The examples show audio URL inputs for lip sync and avatar models.
Can it generate portraits too?
Yes. The source includes a workflow using P-Image before creating the avatar video.
What should I prepare before using it?
Prepare an approved portrait, a short script, voice preferences, output resolution, and any required disclosure text.
Is consent handled automatically?
No. You must verify rights, consent, and disclosure requirements before generating or publishing avatar videos.
Developer Details
Author
halt-catch-fireLicense
MIT
Skillstore revision
r1
Version notice
The author did not declare a version.
Ref
30c73eac2afe762f6aa9c4553158769369d47351
Maintenance freshness
7/18/2026
Usage
12 downloads · 0 views
File structure
📄 SKILL.md
More from halt-catch-fire
View all📦
video-ad-specs
77Create Platform-Ready Video Ads
Teams often create video ads with the wrong dimensions, duration, captions, or safe zones for each platform. This skill gives Claude, Codex, and Claude Code a platform-specific guide for planning video ads and generating CLI-ready creative prompts.
Design & Creative View
📦
product-hunt-launch
77Plan Product Hunt Launches
Product Hunt launches require precise timing, clear positioning, and strong visuals. This skill gives teams a structured checklist for listing copy, gallery assets, comments, research, and launch day execution.
Productivity View
📦
ai-video-generation
75Create AI Videos with Inference.sh Models
Teams need fast ways to turn prompts, images, and audio into video assets. This skill provides model choices and CLI examples for inference.sh video workflows.
Design & Creative View
📦
remotion-render
50Render Remotion Videos From TSX
Programmatic video rendering is difficult when teams need repeatable motion graphics from React components. This skill guides Claude, Codex, or Claude Code to call the inference.sh renderer with Remotion inputs.
Coding & Development View
📦
landing-page-design
50Design Higher-Converting Landing Pages
Landing pages often fail because message, proof, and CTA placement are unclear. This skill gives Claude, Codex, and Claude Code practical rules for hero sections, social proof, forms, mobile layout, and AI-generated visuals.
Design & Creative View
Related Skills
View all📦
p-video
78Generate Pruna Videos from Text and Images
by inferen-sh
Creating video model commands requires choosing compatible inputs and settings. This skill guides Pruna P-Video and WAN generation through the inference.sh CLI.
Design & Creative View
📦
qwen-image-2-pro
79Generate Text-Rich Images with Qwen Pro
by inference-shell
Design teams need image prompts that preserve text and layout details. This skill guides Claude, Codex, or Claude Code to call Qwen-Image-2-Pro through inference.sh.
Design & Creative View
📦
controlnet-pose
79Create Pose-Controlled Images and Videos
by doany-ai
Pose and motion control workflows can be hard to route across image and video models. This skill selects the right RunComfy path and prepares clear CLI guidance.
Design & Creative View
📦
ponyflash
78Generate and Edit AI Media with PonyFlash
by ponyflash
AI media work often spans separate generation, editing, and subtitle tools. PonyFlash combines SDK guidance with FFmpeg workflows for images, videos, speech, music, and exports.
Design & Creative View
📦
image-to-video
78Animate Images with RunComfy Video Models
by runcomfy-com
Still images need the right model and prompt to become useful video clips. This skill routes image-to-video requests to RunComfy models with matching settings and examples.
Design & Creative View
📦
ai-image-generation
78Generate AI Images With Inference.sh
by tool-belt
Teams need fast access to many hosted image models without learning each provider API. This skill guides Claude, Codex, and Claude Code through inference.sh CLI workflows for generation, editing, upscaling, and stitching.
Design & Creative View