Yes. The Wizard maps gf_form, gf_entry, gf_entry_meta, the Slack rows in gf_addon_feed, and gf_addon_feed_log into named variables your system message can use. Variables resolve at request time, so a new notification or a fresh failure shows up in the next answer without re-syncing.

Yes. The feed meta stores the message template with mergetags. SleekAI applies the mergetags against the entry values and returns the rendered message, so you see exactly what landed in the channel, not a sanitized summary. Useful when the template references custom fields.

Yes. The feed records either a channel ID (newer Slack app integrations) or a legacy incoming webhook URL with an embedded default channel. SleekAI exposes either form as a variable and resolves the channel name from the workspace so the bot can say '#sales-leads' rather than C04ABCDXYZ.

Not if you scope it correctly. Use SleekAI's display conditions to require the administrator or editor role, restrict to /wp-admin/ URLs, and require a logged-in state. The widget will not render anywhere else and the data variables will not resolve outside that scope.

Yes. The Slack Add-On stores the API error code (channel_not_found, not_in_channel, invalid_auth, rate_limited) in the feed log. SleekAI reads that row and explains the failure in plain language so the team knows whether to fix the channel ID, re-invite the bot, or wait out a rate limit.

Yes. Multibot mode runs several chatbots on one site. Pair the sales team with one bot scoped to sales feeds and the marketing team with another scoped to their channels. Each gets a different system prompt and a different variable set in the same WP Admin.

The Slack Add-On posts new messages by default. If your feed posts to a thread (via thread_ts) the meta records the parent message reference, and SleekAI surfaces that in answers so the bot can say 'replied in the New Hires thread' rather than treating every post as a fresh top-level message.

You need an account with at least one supported provider. SleekAI works with OpenAI, Anthropic, Google AI Studio, and OpenRouter, so a cheap fast model handles notification lookups while a stronger one writes longer audit reports. Usage and billing live on your provider account.