Agent skill

Siyuan MCP Visual Assets

by yangtaihong59 in yangtaihong59/siyuan-plugins-mcp-sisyphus

MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding.

MITAuto-check passedAI & LLM Engineering

Install Siyuan MCP Visual Assets

skills CLI
$ npx skills add yangtaihong59/siyuan-plugins-mcp-sisyphus --skill siyuan-mcp-visual-assets -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install yangtaihong59/siyuan-plugins-mcp-sisyphus siyuan-mcp-visual-assets --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/yangtaihong59/siyuan-plugins-mcp-sisyphus.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/siyuan-mcp/siyuan-mcp-visual-assets .claude/skills/siyuan-mcp-visual-assets && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
siyuan-mcp-visual-assets
GitHub stars
116
Token cost
~1.1k tokens
SKILL.md length
487 words
Files
2
Skills in repo
30
Repo updated
First seen
Licence
MIT

At a glance

MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding.

  • Works in 4 steps: Read the source image or data and record… → Reuse a matching deterministic… → Check escaped text, viewBox containment,… → …
  • Tasks that involve MCP servers
  • SKILL.md covers Generate and review, Upload and embed and Structural readback
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Siyuan MCP Visual Assets is an agent skill from yangtaihong59/siyuan-plugins-mcp-sisyphus. MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding.

Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).

It sits in AI & LLM Engineering, covering MCP servers and Embeddings. It works with Model Context Protocol. The repository describes itself as: 一个用于连接Agent与Siyuan的工具. The licence is MIT.

When your agent uses it

  • Tasks that involve MCP servers
  • Tasks that involve Embeddings

Example prompts

  • “/siyuan-mcp-visual-assets”

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. Read the source image or data and record chart type, series, categories, values, axes, units, grid lines, markers, title, legend…
  2. Reuse a matching deterministic config/builder shape for charts; for figures, decode structure first and choose semantic geometry. If a…
  3. Check escaped text, viewBox containment, text bounds and collisions, line/text clearance, stable stroke width, angle/projection/occlusion…
  4. Semantic correctness comes before pixel similarity. Pixel overlay is optional. A real SiYuan UI review is required for display claims…

What it can do on your machine

Read from SKILL.md and the folder at commit d2ad534. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Siyuan MCP Visual Assets loads about 1.1k tokens when it runs. Until then it costs about 34 tokens; SKILL.md has 487 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~34
When it runs · the whole SKILL.md, loaded when a task matches
~1.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from yangtaihong59/siyuan-plugins-mcp-sisyphus at commit d2ad534, republished under its MIT licence (© yangtaihong59). 487 words, ~1,108 tokens.

Download SKILL.mdSave it as .claude/skills/siyuan-mcp-visual-assets/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
siyuan-mcp-visual-assets
description
MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding.

SiYuan Visual Assets with MCP

Choose the narrowest route: data/config-driven SVG for statistical charts, semantic geometry for explanatory figures, and an existing uploaded asset when vector reconstruction is not appropriate. This scenario does not add a picture-generation tool, renderer, async runner, or platform-specific image utility.

Generate and review

  1. Read the source image or data and record chart type, series, categories, values, axes, units, grid lines, markers, title, legend, geometry, and unresolved readings. Do not invent unclear values.
  2. Reuse a matching deterministic config/builder shape for charts; for figures, decode structure first and choose semantic geometry. If a supported mathematical relation exists, declare and run a validation such as growth or sum; failed validation returns to data review.
  3. Check escaped text, viewBox containment, text bounds and collisions, line/text clearance, stable stroke width, angle/projection/occlusion constraints, and one unique data-chart-key matched to the config registry.
  4. Semantic correctness comes before pixel similarity. Pixel overlay is optional. A real SiYuan UI review is required for display claims: width, responsive behavior, readability, theme contrast, and embed rendering.

Do not make bitmap tracing the default route. Use it only as constrained coordinate assistance; do not force-vectorize photos, maps, comics, or figures whose meaning is the typeface itself. If text, a list, table, or formula states the meaning clearly, prefer native content.

Upload and embed

Resolve the exact notebook/document/block ID, parent-child relation, and insertion position before writing. A title, search hit, or UI position is not write authorization.

text
file(action="upload_asset", assetsDirPath="<approved-assets-dir>", localFilePath="<selected-local-file>")

Asset upload reads a user-selected local file and requires explicit approval of both the source file and assets directory. Use the returned asset path as the only subsequent reference; never guess a timestamped filename or preserve a machine-specific path in the Skill.

For an ordinary uploaded image, append one bounded Markdown image block. For inline SVG/HTML, keep all lines in one standalone block; use a complete NodeHTMLBlock DOM update only when an existing block is the exact target. A canonical visual block may be reused elsewhere with a query embed pointing to its stable block ID; do not duplicate the SVG body.

text
block(action="append", parentID="<resolved-parent-id>", dataType="markdown", data="<one-block-markdown-or-inline-html-svg>")
text
block(action="update", items=[{"id":"<existing-leaf-block-id>","dataType":"dom","data":"<complete-node-html-block-dom>"}])
Show full SKILL.md (142 more words)Show less

Structural readback

After each write, read the exact affected ID before any search or UI conclusion:

text
block(action="get_kramdown", id="<written-block-id>")
text
block(action="get_children", id="<resolved-parent-id>", page=1, pageSize=200)
text
block(action="dom", id="<written-block-id>")
text
file(action="get_doc_assets", id="<resolved-document-id>", assetType="all")
text
search(action="search_assets", query="<asset-filename>", exts=["svg","png"])
text
file(action="read_image", id="<resolved-document-id>", path="assets/<returned-image-path>")

Confirm block type, parent/sibling order, kramdown, DOM attributes, viewBox, escaped text, returned asset path, data-chart-key, canonical embed target, and image/HTML containment. get_doc_assets and search_assets find references or candidates; neither proves rendering. For a bitmap already referenced by the authorized document, a vision-capable client may call read_image for that one image; it still does not replace a real SiYuan UI rendering review. API success, file existence, SQL/index results, or a renderer invocation cannot replace structural readback or real SiYuan UI review.

If a response is lost or the returned asset path is missing/non-unique, stop and inspect the exact target; do not retry an upload or append blindly. Classify geometry WARNs as repaired, justified exemption, or TODO; after repeated unsuccessful adjustments, stop and report instead of relaxing tolerances.

© yangtaihong59, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in skills/siyuan-mcp/siyuan-mcp-visual-assets of yangtaihong59/siyuan-plugins-mcp-sisyphus.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit d2ad534

Compare with similar skills

Siyuan MCP Visual Assets next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Siyuan MCP Visual Assets compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Siyuan MCP Visual Assets this skillyangtaihong59/siyuan-plugins-mcp-sisyphus116—~1.1kAutomated safety check: PassMIT
Copilot SDKgithub/awesome-copilot40k4 repos~6.3kAutomated safety check: PassMIT
Pi AgentK-Dense-AI/scientific-agent-skills48k1 repos~2.1kAutomated safety check: PassMIT
Sandbaseiflytek/skillhub5.2k1 repos~2.1kAutomated safety check: PassApache-2.0
Sandbasesandbaseai/cli188—~3.1kAutomated safety check: PassApache-2.0
Cookbook Aimldatabricks-solutions/databricks-apps-cookbook183—~1.7kAutomated safety check: PassCustom licence

Similar skills

  • Copilot SDK

    github/awesome-copilot

    Official

    Build agentic applications with GitHub Copilot SDK. An agent skill from github/awesome-copilot.

    40k GitHub starsUsed in 4 repos~6.3k tokens
    AI & LLM EngineeringAuto-check passed
  • Pi Agent

    K-Dense-AI/scientific-agent-skills

    Builds with and operates Pi, the minimal terminal coding harness.

    48k GitHub starsUsed in 1 repo~2.1k tokens
    AI & LLM EngineeringAuto-check passed
  • Sandbase

    iflytek/skillhub

    Access 2,000+ AI models and API tools through one MCP interface for inference, media generation, search, scraping, embeddings, social data, and structured retrieval.

    5.2k GitHub starsUsed in 1 repo~2.1k tokens
    AI & LLM EngineeringAuto-check passed
  • Sandbase

    sandbaseai/cli

    Discover better ways to complete tasks. An agent skill from sandbaseai/cli.

    188 GitHub stars~3.1k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Cookbook Aiml

    databricks-solutions/databricks-apps-cookbook

    Invoke ML models, run vector search, and connect to MCP servers from Databricks Apps.

    183 GitHub stars~1.7k tokensUpdated 6 days ago
    AI & LLM EngineeringAuto-check passed
  • Civitai

    artokun/comfyui-mcp

    Discover Civitai models with the BUILT-IN downloadmodel action:"searchcivitai" and install/generate them locally.

    803 GitHub stars~1.1k tokensUpdated 6 days ago
    AI & LLM EngineeringAuto-check passed

More from yangtaihong59/siyuan-plugins-mcp-sisyphus

All 30 skills in this repo
  • Siyuan Built Cjs Live Test

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    使用仓库刚构建的 dist/mcp-server.cjs 与 cli/dist/cli.cjs,在真实思源实例的隔离测试笔记本中执行端到端验收。用于验证工具或 action 修改、严格安全写入、短哈希预检租约、幂等重放、状态变化拒绝、HTTP/stdio/CLI 一致性,以及用户要求“实际测试”“真实笔记本测试”“全部 action 测试”或“不要只跑 mock”时。

    116 GitHub stars~1.3k tokensUpdated 3 days ago
    Auto-check passed
  • Claude Desktop MCP App

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    Configure, repair, and verify MCP servers and MCP Apps in Claude Desktop on macOS.

    116 GitHub stars~1.2k tokensUpdated 3 days ago
    Auto-check passed
  • Git Workflow Cn

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    Chinese git workflow conventions including commit message generation and merge practices.

    116 GitHub stars~941 tokensUpdated 3 days ago
    Auto-check passed
  • Siyuan MCP Browse Read

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    MCP playbook for browsing and reading SiYuan notes. An agent skill from yangtaihong59/siyuan-plugins-mcp-sisyphus.

    116 GitHub stars~684 tokensUpdated 3 days ago
    Auto-check passed
  • Siyuan MCP Create Edit

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    MCP playbook for creating and editing SiYuan documents and blocks.

    116 GitHub stars~665 tokensUpdated 3 days ago
    Auto-check passed
  • Siyuan MCP Sisyphus

    yangtaihong59/siyuan-plugins-mcp-sisyphus

    Top-level skill for operating SiYuan Note through the Sisyphus MCP server.

    116 GitHub stars~1.3k tokensUpdated 3 days ago
    Auto-check passed

Questions about Siyuan MCP Visual Assets

What does Siyuan MCP Visual Assets do?

MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding. Siyuan MCP Visual Assets is an agent skill from yangtaihong59/siyuan-plugins-mcp-sisyphus. MCP playbook for semantic SVG charts and figures, approved asset upload, and verified SiYuan block embedding.

When should I use Siyuan MCP Visual Assets?

Siyuan MCP Visual Assets fits situations like: tasks that involve MCP servers; tasks that involve Embeddings.

How do I install Siyuan MCP Visual Assets in Claude Code?

Run `npx skills add yangtaihong59/siyuan-plugins-mcp-sisyphus --skill siyuan-mcp-visual-assets -a claude-code`. Or copy the skill folder (skills/siyuan-mcp/siyuan-mcp-visual-assets in yangtaihong59/siyuan-plugins-mcp-sisyphus) into .claude/skills/siyuan-mcp-visual-assets in your project. Claude Code loads it when a task matches its description.

How do I install Siyuan MCP Visual Assets in Codex?

Run `npx skills add yangtaihong59/siyuan-plugins-mcp-sisyphus --skill siyuan-mcp-visual-assets -a codex`. Or copy the skill folder (skills/siyuan-mcp/siyuan-mcp-visual-assets in yangtaihong59/siyuan-plugins-mcp-sisyphus) into .agents/skills/siyuan-mcp-visual-assets in your project. Codex loads it when a task matches its description.

Can I use Siyuan MCP Visual Assets in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add yangtaihong59/siyuan-plugins-mcp-sisyphus --skill siyuan-mcp-visual-assets -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/siyuan-mcp-visual-assets, .gemini/skills/siyuan-mcp-visual-assets, .github/skills/siyuan-mcp-visual-assets and .opencode/skills/siyuan-mcp-visual-assets in your project.

What does Siyuan MCP Visual Assets need to run?

SKILL.md names no scripts, command-line tools or credentials: Siyuan MCP Visual Assets is instructions for the agent only.

Does Siyuan MCP Visual Assets access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Siyuan MCP Visual Assets safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Siyuan MCP Visual Assets use?

Siyuan MCP Visual Assets is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Siyuan MCP Visual Assets use?

About 1.1k tokens (SKILL.md is roughly 4.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Siyuan MCP Visual Assets?

Skills that share tags, products or a category with Siyuan MCP Visual Assets: Copilot SDK (github/awesome-copilot, 40k stars), Pi Agent (K-Dense-AI/scientific-agent-skills, 48k stars), Sandbase (iflytek/skillhub, 5.2k stars) and Sandbase (sandbaseai/cli, 188 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Siyuan MCP Visual Assets?

yangtaihong59 (a GitHub user) maintains it in yangtaihong59/siyuan-plugins-mcp-sisyphus, which has 116 GitHub stars. The repository holds 30 skills in this directory. The repository was last updated on October 8, 2026.

Source: yangtaihong59/siyuan-plugins-mcp-sisyphus on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.