# Changelog

## Score

- CAI 53 → 52 (-0.3)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.2) — scores are not directly comparable.

## Lenses

- Code Health 62 → 62 (+0.0)
- Architecture 98 → 98 (+0.0)
- Maturity 57 → 60 (+3.4)
- Readiness 67 → 67 (+0.0)
- Security 41 → 38 (-2.5)

## Resolved (6)

- Ambiguous naming for QR code creation. `qrcode_create_scene` and `qrcode_create_limit_scene` likely correspond to WeChat's 'long-term' vs 'short-term' QR codes, but the names are not intuitive. `limit` is not a standard term for 'short-term' or 'expire'.
- Duplicate intent across Api and MpApi. Both `Api` and `MpApi` have `template_message_send`. Since `MpApi` is likely a subclass or specific implementation for Official Accounts, having the same method in both suggests either poor inheritance usage or redundant API surface.
- Inconsistent entry points for obtaining API instances/config. `Wechat.api` returns an API instance, `ApiLoader.config` returns config, and `ControllerApi.wechat` returns a responder/api wrapper. The naming `api` vs `wechat` vs `config` is inconsistent across modules for similar intents (getting the API client).
- Inconsistent naming for message sending operations. `message_mass_sendall` uses 'mass' and 'sendall', while `custom_message_send` uses 'custom' and 'send'. The verb 'send' vs 'sendall' and noun 'mass' vs 'custom' creates inconsistency in how different message types are addressed.
- Off-boarding risk: anonymized user #1
- Redundant/Confusing naming in WechatResponder. `wechat_api` and `wechat_responder` appear to be aliases or near-duplicates for configuring the responder, while `wechat` takes an account. The distinction between 'api' and 'responder' in the method names is unclear and likely redundant given the class context.

## New (1)

- Off-boarding risk: anonymized user #1
