Feature: docs +fetch should support --resolve-media to auto-download images/files

[liuhedev]

Problem

When using docs +fetch to export a Lark document as Markdown, images and files are exported as placeholder tags:

<image token="DKblbz86oojD..." width="1833" height="2491"/>
<file token="WIKKbg01lonJ..." name="skills.zip"/>

These placeholders cannot be rendered in any Markdown viewer. Users must manually call docs +media-download for each token to get the actual files.

Critical issue: When fetching documents from shared Wiki spaces, docs +fetch has read permission at fetch time, but docs +media-download called later often returns HTTP 403 Forbidden — the media download API requires separate permission that may not be available after the initial fetch.

Proposed Solution

Add a --resolve-media flag to docs +fetch that:

1. During the fetch, identifies all <image>, <file>, and <whiteboard> placeholder tags
2. Immediately downloads each media asset using the same auth context (while permission is still valid)
3. Saves assets to a local directory (e.g., assets/<doc-name>/)
4. Replaces placeholder tags with standard Markdown references (![](assets/...), [file](assets/...))

# Current behavior (placeholders only)
lark-cli docs +fetch --doc /wiki/TOKEN

# Proposed behavior (with media resolution)
lark-cli docs +fetch --doc /wiki/TOKEN --resolve-media
lark-cli docs +fetch --doc /wiki/TOKEN --resolve-media --media-dir ./images

Workaround

I wrote a Python script that does this post-hoc, but it fails with 403 on shared documents because the media download permission window has closed: https://gist.github.com/liuhedev

Environment

• lark-cli version: latest
• OS: macOS arm64
• Auth: user identity (feishu)