onboarding: sweep every known chat, so no group is ever left behind by hand

onboarding: never let a network blip trigger a leave + blocklist

入群授权催办：没权限就不判定，10 分钟仍未授权自动退群拉黑

可选的跨供应商兜底：主供应商超时不再直接掉到只认关键词的词库

封版审计修复：健康端口不再被卡死 + token 停止进日志

反spam: 同步私有姊妹版的入口完整性审计成果

反spam: 堵住五个静默绕过 + 新增签到探针层

i18n: match against full navigator.languages preference list

i18n: internationalize landing page + READMEs (EN·中文·RU·ES·AR·FA)

HF Space: fix sdk static->docker so a fork actually runs the bot

Brand: lead READMEs with antispam.bot (TelegramGuard as project name)

Tidy: drop vestigial device field + rewrite READMEs for the flagship

Phase 3 clean-up: purge grey-market persona + fix broken setup docs

Phase 3: 落地页隔离进 site/ 子目录(CF Pages 输出目录锁定 site,不暴露 bot 源码)

Phase 3: 落地页换成设计师版 antispam.bot(纯静态单文件,CF Pages 就绪)

Phase 2: 中文黑话词库 + I-Lang v5.0 判定参考 + 修子串误判

Phase 1: 去 Gemini，AI 改走 OpenAI 兼容 provider（默认硅基流动）

Eastsoft Inc. -> iLang Inc.

fix: webhook mode starts listening immediately, no AI test blocking

v4.5: HF Space → static promo page + Cloud Run as recommended deploy

v4.5: HF Space → static landing page, bot runs on Cloud Run

v4.4: webhook mode for Cloud Run / Railway / Render

fix: timeout params on builder not run_polling

fix: bootstrap_retries=10 + 30s timeouts for HF Space network

fix: increase Telegram API timeouts for HF Space network

fix: DB_PATH default to /data/bot.db for HF persistent storage

v4.3: pre-filter layer — zero API cost spam defense

add MIT LICENSE file

v4.2: README rewrite — user journey (use now → install → deploy) + Chinese version

v4.1: one-line installer + README redesign

fix: use codename ACC, not internal nickname

v4.0: SOUL+PUBLIC+PROPAGATION — 骚货's cognitive engine + three-step method + I-Lang evangelism

v3.2: THREE_STEP method for risky topics — insider risks → reality check → direction

fix: pass BLOCK_NONE safety_settings on every API call, not just constructor

v3.1: ZERO_MORALIZE gene — never lecture, never disclaim, just answer

fix: pin google-generativeai==0.8.4 + startup AI test

fix: safe Gemini response extraction + startup logging

v3.0: international edition — universal prompts, English UI, multilingual bot

fix: relax Gemini safety filters — BLOCK_NONE for all categories

v2.0: full-power prompts — 雅痞人格+黑话词典+看图三步法+三步反spam

fix: colorTo must be from HF allowed list

fix: add sdk:docker to HF Space config

fix: HF Space health check on port 7860 + Node.js 24 Actions warning

add HF metadata + fix warnings

add datasets sync workflow

v1.0: TelegramGuard - AI-powered group guardian with I-Lang Prompt Spec