OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Workflow Agents

Hbg Classical Poem Silk Video

Hbg Classical Poem Silk Video is an installable agent skill that guides a workflow for turning Chinese classical poems into vertical Chinese-art videos, from poem analysis and scene planning through image-to-video generation, captions, audio and final MP4 checks.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Hbg Classical Poem Silk Video — OpenViST project profile in Workflow Agents.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want an agent to guide a structured Chinese-poem video workflow, or when reviewing generated clips and an exported MP4 against the skill’s stated checks.

Who it is for

  • Creators making illustrated vertical videos from Chinese classical poetry.
  • Codex or Claude Code users who want an agent-guided poem-to-video workflow.
  • Video creators reviewing generated clips for visual or encoding issues.

Documented uses

  • Break a poem into scenes based on lines, imagery and emotional progression.
  • Generate illustrated stills and animate selected elements while keeping other composition elements stable.
  • Add brush-style captions, retain generated ambience, add background music and transition between scenes.
  • Inspect an encoded MP4 for visual and media-quality issues.
Before you choose

Requirements & limits

The repository is public and lists an MIT License. Installation instructions cover Codex and Claude Code; the README also mentions agents that support SKILL.md. The documented image-to-video stage depends on a separate Docker runtime and user-controlled authorization.

  • Still-image generation uses the agent’s built-in ImageGen; image-to-video generation is delegated to the separate hbg-gemini-flow-suite Docker runtime.
  • The README specifies a user-authorized runtime setup before running the skill’s prerequisite check.
  • The skill’s workflow uses fixed framing, visual anchors and limited scene actions to constrain image-to-video changes.
  • Final review is performed on the encoded MP4, including frame extraction and checks for black frames, transitions, audio and other listed issues.
  • The repository lists an MIT License for the skill, scripts and documentation; the included Ma Shan Zheng font is identified separately under the SIL Open Font License.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the poem-to-video skill in Codex or Claude Code

Install and check the classical-poem-silk-video agent skill before using it.

Before you start

  • Codex or Claude Code, or another agent that supports SKILL.md, for the natural-language installation route.
  • For the CLI route, a shell with curl available.

Steps

  1. Choose an installation route: ask a supported agent to install the skill from the repository, or use the README’s Codex or Claude Code install command.

  2. For manual installation, clone the repository, create the Codex skill directory, and copy the skill contents into it.

  3. Check that SKILL.md, agents, references, scripts, and font assets are present; the README’s agent-install instructions also request shell syntax, Python compilation, and skill validation checks.

Expected outcome

The skill is installed in the selected agent’s skills location and its files and checks can be reviewed before use.

Guide limitations

  • The documented automatic checks are requested as part of the agent installation prompt; the README does not provide their individual commands or outputs.
  • Image-to-video generation and optional watermark cleanup require the separate hbg-gemini-flow-suite Docker runtime and user-controlled authorization; these are not part of installing the skill.
Documentation 1 ↗
Source-based practical guide

Run final media checks on an exported MP4

Generate QA reports and sample frames from an exported video using the skill’s final-media QA script.

Before you start

  • An exported MP4 file.
  • The installed skill script at skill/classical-poem-silk-video/scripts/final_media_qa.sh.

Steps

  1. Run the script with the MP4 path and a QA output directory: skill/classical-poem-silk-video/scripts/final_media_qa.sh final.mp4 qa/final.

  2. Review the generated ffprobe, black-detection, silence-detection, and volume-detection reports, along with sampled frames at 10%, 25%, 50%, 75%, 90%, and the end.

  3. For targeted checks, prepare a TSV of timestamps and labels, then rerun the script with that TSV as the third argument.

Expected outcome

The QA directory contains media-analysis reports and sampled frames from the encoded video; an optional timestamp TSV adds selected samples such as transition midpoints.

Guide limitations

  • This procedure checks an existing MP4; it does not describe how to produce or repair one.
  • The README lists outputs but does not specify pass/fail thresholds or guarantee that the video will pass QA.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
Mr-funny/hbg-classical-poem-silk-video · Official homepage title · retained in its source language
How we review sourcesSuggest a correction