4 Commits

Author	SHA1	Message	Date
kappa	1ebe765691	Replace pattern-based agent routing with AI intent classification ... Instead of regex patterns to detect user intent, use a lightweight AI call (classifyIntent) that categorizes messages into troubleshoot, onboarding, billing, asset, or general. This catches ambiguous expressions like "example.com 좀 봐주세요" that patterns would miss. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-12 02:38:55 +09:00
kappa	0e338aa0fa	Fix feedback prompt by returning agent name from route registry ... routeToActiveAgent now returns {response, agentName} so the message handler can reliably determine which agent handled the request. Feedback is only prompted for troubleshoot and onboarding agents, not billing or asset single-shot agents. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-11 20:42:25 +09:00
kappa	42e7f4a72a	Only prompt feedback after multi-round agent sessions ... Billing and asset agents (single-shot) no longer trigger the star rating feedback prompt after every response. Feedback is now only requested after troubleshoot/onboarding sessions or when the conversation had 3+ exchanges in the last hour. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-11 20:38:35 +09:00
kappa	1d6b64c9e4	Initial implementation of Telegram AI customer support bot ... Cloudflare Workers + Hono + D1 + KV + R2 stack with 4 specialized AI agents (onboarding, troubleshoot, asset, billing), OpenAI function calling with 7 tool definitions, human escalation, pending action approval workflow, feedback collection, audit logging, i18n (ko/en), and Workers AI fallback. 43 source files, 45 tests passing. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-02-11 13:21:38 +09:00