tkmind/memind
1
0
Fork 0
Code Issues Pull Requests 2 Actions Packages Projects Releases Wiki Activity

feat(wechat): add WeChat voice reco API fallback for service account
Memind CI / Test, build, and release guards (push) Successful in 3m27s
Details

Browse Source 
When passive Recognition is empty, convert AMR to mp3 and call WeChat
addvoicetorecofortext before the existing asr.tkmind.cn fallback.

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
john
2026-08-05 18:01:50 +08:00
parent d9db72fd90
commit 26846a0274
7 changed files with 529 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.env.example
+5
View File
@@ -338,6 +338,11 @@ H5_ACCESS_PASSWORD=change-me
# H5_ASR_TARGET=https://asr.tkmind.cn
# H5_ASR_MAX_BYTES=5242880
# H5_ASR_TIMEOUT_MS=45000
# 微信服务号语音：Recognition 为空时优先走微信 addvoicetorecofortext（需 ffmpeg 转 mp3）；失败再回落 H5_ASR
# H5_WECHAT_MP_VOICE_RECO_API=1
# H5_WECHAT_MP_VOICE_RECO_LANG=zh_CN
# H5_WECHAT_MP_VOICE_RECO_API_BASE=https://api.weixin.qq.com
# H5_FFMPEG_PATH=ffmpeg
# 前端构建时注入（Vite，需 VITE_ 前缀）
# 工作目录：新建会话时使用，必填