labsai/EDDI

Problem / Motivation WhatsApp (2B+ users globally) is the dominant messaging platform for customer engagement in APAC, LATAM, Europe, and Africa. EDDI's conversational agents can power customer support, sales qualification, and onboarding workflows via WhatsApp Business. Note: WhatsApp Business API is primarily 1:1 (business-to-customer). A Groups API exists for Official Business Accounts (max 8 participants), but it is a restricted, premium feature - not the standard use case. The initial adapter should support AGENT targets only. GROUP target support via the Groups API can be a future enhancement. Key WhatsApp constraints: Primarily 1:1 - Standard use case is business-to-customer direct messaging Groups API - Exists but requires Official Business Account (OBA) status, max 8 participants 24-hour messaging window - Businesses can only send free-form messages within 24 hours of the customer's last message. Outside this window, only pre-approved message templates are allowed. Cloud AP

Problem / Motivation Microsoft Teams (320M+ monthly active users) is the dominant enterprise communication platform. EDDI's multi-agent orchestration capabilities - especially group discussions with debate, peer review, and round table styles - are a natural fit for enterprise Teams channels. WARNING: Teams has the most complex authentication model of all target platforms. The Bot Framework uses Azure AD JWT tokens for incoming message verification, requiring OpenID metadata fetching, key rotation handling, and multi-claim validation. Plan for this complexity. Key Teams constraints: No threading in Group Chats - Threads (reply chains) are only supported in Channels. Group Chats are flat message streams. Multi-agent debates MUST deploy in Channels. @mention-gated by default - Bots only receive messages when explicitly @mentioned unless Resource-Specific Consent (RSC) is configured with ChannelMessage.Read.Group permission. Adaptive Cards (v1.5) - Teams uses Adaptive Cards for rich fo

Problem / Motivation Discord (200M+ monthly active users) is a major platform for developer communities, open-source projects, and increasingly enterprise teams. EDDI's multi-agent group discussions map naturally to Discord's channel + thread model. WARNING: Critical architectural difference from Slack: Discord's HTTP-based Interactions API only receives slash commands and component interactions (button clicks, select menus). Regular channel messages and thread replies require a persistent WebSocket Gateway connection. This fundamentally differs from Slack's pure HTTP Events API model. The recommended approach is a phased implementation. Architecture Principle No platform SDKs. All adapters use java.net.http.HttpClient + Jackson for raw HTTP/JSON - no JDA, no Discord4J. Every messaging platform's API is just REST under the hood. Raw HTTP keeps the dependency tree lean, the single-JAR deployment simple, and gives full control over retry logic, error handling, and message formatting.

Problem / Motivation Telegram (950M+ monthly active users) is the architecturally closest platform to Slack for EDDI integration: HTTP webhook model - Telegram sends POST requests to a registered URL (identical to Slack's Events API) Simple secret verification - String comparison of X-Telegram-Bot-Api-Secret-Token header (simpler than Slack's HMAC-SHA256) Simple REST API - sendMessage, sendChatAction with JSON bodies Reply threading - reply_to_message_id creates visual reply chains Generous message limit - 4096 characters (vs Slack's ~4000) No OAuth complexity - Single bot token, no workspace/tenant scoping This makes Telegram the recommended first non-Slack adapter. Architecture Principle No platform SDKs. All adapters use java.net.http.HttpClient + Jackson for raw HTTP/JSON. No TelegramBots library. Every messaging platform's API is just REST under the hood. Raw HTTP keeps the dependency tree lean, the single-JAR deployment simple, and gives full control over retry logic, error ha

Summary Three REST interfaces are missing core OpenAPI annotations that all other REST endpoints have. What to add Interface File What's missing IRestUserConversationStore engine/triggermanagement/IRestUserConversationStore.java @Operation on all 3 methods IRestHtmlChatResource ui/IRestHtmlChatResource.java @Operation on both methods ILogoutEndpoint engine/api/ILogoutEndpoint.java @Tag class-level annotation (methods already have @Operation) Pattern @Operation — add a summary field: @Operation(summary = "Read user conversation", description = "...") @Tag: @Tag(name = "Authentication") Tags should match an entry in src/main/java/ai/labs/eddi/configs/OpenApiConfig.java. If no matching tag exists, add one there too. Verify before starting: Confirm these annotations are still missing — the codebase may have been updated since this issue was filed. Acceptance criteria IRestUserConversationStore: all 3 methods have @Operation IRestHtmlChatResource: both methods have @O