What

Improve Discord and Telegram connect error handling so users see accurate, localized failure messages and diagnostics include enough proxy/runtime context to debug real connection issues.

Why

Discord and Telegram connect failures were too opaque during user debugging:

• the frontend could show misleading success-style messaging for connect failures
• backend responses flattened distinct failure modes into generic connect errors
• failure logs were missing the proxy/runtime context needed to quickly identify network and proxy issues

How

• add structured channel connect error responses with code, phase, retryable, and requestId
• classify Discord/Telegram failures into credential, app-id mismatch, timeout, network, persist, and sync phases
• localize backend error descriptions from controller locale instead of surfacing raw internal errors
• update web to show returned error messages for Telegram/Discord and stop treating generic connect failures as already-connected states
• log proxy/runtime context on connect failures and bound async runtime health capture with a short timeout
• regenerate OpenAPI and web SDK types for the new error contract

Affected areas

• Desktop app (Electron shell)
• Controller (backend / API)
• Web dashboard (React UI)
• OpenClaw runtime
• Skills
• Shared schemas / packages
• Build / CI / Tooling

Checklist

• pnpm typecheck passes
• pnpm lint passes
• pnpm test passes
• pnpm generate-types run (if API routes/schemas changed)
• No credentials or tokens in code or logs
• No any types introduced (use unknown with narrowing)

Notes for reviewers

pnpm test hit an unrelated flaky failure in tests/desktop/skill-dir-watcher-workspace.test.ts (Timed out waiting for watcher state while waiting for agent-tool to appear for bot-1).