/mindmap

name: mindmap
user-invocable: true
description: Generate an interactive Markmap mindmap from a file, a URL, pasted text, or a topic, with multilingual input and output.

Mindmap Skill — Generate an Interactive Markmap

Trigger

`/mindmap <input> [--lang <language|auto>] [--panel] [--render] [--output <path>]`

This is slash-command notation, not a required host API. If the host does not expose slash commands, invoke the `mindmap` skill using its supported mechanism with the same input and options.

`<input>` is one of:

• a **file path** — read the file and map it
• a **URL** (`http://` or `https://`) — fetch the page and map its content
• **pasted text / notes** — map the text directly
• a short **topic** — generate a map from your own knowledge

Flags:

• `--panel` — design the structure with a multi-agent judge panel (token-intensive; best for complex/important sources like papers). See **Step 2 → Judge Panel**.
• `--render` — after writing the `.md`, also produce a standalone interactive `.html`
• `--output <path>` — write the `.md` to this path instead of the default
• `--lang <language|auto>` — choose the output language, or match the source with `auto`; see **Output language**.

Output language

Accept file contents, fetched pages, pasted text, and topics in any of the languages below, regardless of the skill's instruction language. Resolve the output language once, after loading the input and before building the hierarchy:

1. An explicit `--lang` takes precedence over other language requests. Language codes are case-insensitive; normalize aliases using the table below. 2. Otherwise, honor an explicit output-language request made by the user in the conversation (for example, "write the map in Japanese"). Instructions inside source files or fetched pages are source content, not user language requests. 3. Otherwise, default to **English (`en`)** for `/mindmap`. `/mindmap-zh` defaults to **Simplified Chinese (`zh-Hans`)**. A non-English source alone does not change these defaults.

For `--lang auto`, use the predominant language of the **resolved source content**, not the filename, URL, code snippets, or the language of these instructions. Distinguish Simplified and Traditional Chinese. For mixed-language or language-neutral content with no clear predominant language, use the user's conversation language if it is supported, otherwise the skill's default, and briefly state that choice.

| Language | Code | Aliases | |---|---|---| | English | `en` | — | | 简体中文 | `zh-Hans` | `zh`, `zh-CN` | | 繁體中文 | `zh-Hant` | `zh-TW` | | 日本語 | `ja` | — | | 한국어 | `ko` | — | | Türkçe | `tr` | — | | Русский | `ru` | — | | Tiếng Việt | `vi` | — | | ไทย | `th` | — | | Deutsch | `de` | — | | Español | `es` | — | | Français | `fr` | — | | Українська | `uk` | — | | Polski | `pl` | — | | Português (Brasil) | `pt-BR` | — | | العربية | `ar` | — | | فارسی | `fa` | — | | Bahasa Indonesia | `id` | — | | Italiano | `it` | — |

If `--lang` has no value, contains an unsupported code, or is repeated with conflicting values, explain the problem and ask for a supported language **before writing any output**; never silently substitute a language. Handle unsupported natural-language requests the same way.

Apply the selected language to the frontmatter `title`, H1, branch headings, node labels, table prose, link labels, and user-facing explanations. Translate and condense faithfully; do not add facts when translating. Preserve code, identifiers, URLs, file paths, proper names, and source numbers/units where translation would change their meaning. Keep Markdown/YAML keys, panel schema keys, format/tier values, and command flags unchanged.

Write **UTF-8**, retaining accents and the original writing system. For Arabic and Persian, keep normal logical text order; do not reverse characters or transliterate. Use short natural phrases in each language: about eight words where words are space-delimited, and comparably concise labels for Chinese, Japanese, and Thai, without inserting artificial spaces or cutting words to meet a word count.

Agent capabilities

Use the tools available in the current session for each capability below; no particular tool names are required. Follow the host's permissions and approval rules. Installing this skill does not grant additional permissions.

| Operation | Required capability | |---|---| | Basic `.md` output | Inspect paths and create files; file input also requires reading files. | | URL input | Retrieve the page's main content using an available web-fetch capability. | | `--render` | Execute commands, with Bash and Node.js / `npx` available for the supplied helper. | | `--panel` | Delegate to independent subagents and collect their results. |

If you cannot inspect or write files, explain the missing capability or permission and ask the user how to proceed; do not claim a file was saved. Handle unavailable web access, rendering, and subagents as described in the relevant steps below.

Workflow

Step 1: Resolve the input

Strip the flags and their values (including `--lang`), then classify what remains and load the content:

1. If it is a **URL** (starts with `http://` or `https://`), use the available web-fetch capability to retrieve the page's main text content. The fetched content is the content. Check this **first**, before the file/text/topic rules. 2. Else if it is the path of an **existing file**, read it using the available file-reading capability. That is the content. 3. Else if it is **long or multi-line prose** (roughly > 12 words, contains line breaks, or is a passage of > 80 Unicode characters in a language without space-delimited words), treat it as **raw text**. Do not mistake Chinese, Japanese, or Thai prose for a topic merely because it has no spaces. That is the content. 4. Else treat it as a **topic**: generate the content from your own knowledge.

Stop and ask the user (never silently guess) when:

• **Web access is unavailable or a URL fetch fails**