feat: hint Claude to read image files via Read tool

[clairewangjia]

Summary

When users send images via WeChat, they are correctly downloaded, decrypted, and saved locally. However, the message passed to Claude Code via notifications/claude/channel only contains a plain text line like:

[图片] 已保存到 /path/to/img_xxx.jpg

Claude sees this as a file path string but has no cue to actually view the image. Since Claude Code's Read tool supports reading image files (it's a multimodal LLM), adding a simple hint resolves the issue:

[图片] 已保存到 /path/to/img_xxx.jpg
请用 Read 工具查看该图片文件来了解图片内容。

This is consistent with how video frames are already listed with paths for Claude to read (lines 385-401 in media.ts).

Changes

• media.ts: In extractContent(), after saving image path to message parts, add a hint telling Claude to use its Read tool to view the image file.

Test plan

• Send an image via WeChat to Claude Code session
• Verify Claude reads the image file and describes its content correctly
• Verify no regression for text messages, voice, video, file handling

🤖 Generated with Claude Code

[LinekForge]

Merged in 32f4657. Thanks for the clean, focused contribution! 🙏

Quick note on why this matters beyond "feature polish": the README already promised 图片 | CDN 下载 + AES 解密，Claude 可查看 — but extractContent was only pushing the file path as plain text, so Claude Code had no cue to actually Read the image. Your PR fixes a docs-code drift, not just adds a feature. That's exactly the right kind of maintainer-helpful contribution.

Also mirroring the video branch (which already enumerates frame paths for Read) — same idea, applied to image. 对偶干净。

Apologies for the 11-day delay — we were heads-down shipping forge-hub v0.2.0 and I missed the notification. Working on tightening our contributor-signal loop now.

Again, thanks. Looking forward to more from you 👋

— Forge (maintainer)