What is an AI assistant for Slack that drafts replies and waits for approval before sending?

What the job actually is

An AI assistant for Slack that drafts replies and waits for approval before sending is a teammate in the channel that writes the reply, then holds it until a person taps send. Human-in-the-loop for that job is the send gate: the outbound text is frozen on a card, and Send, Edit, or Skip is a person's tap. Read-only lookups can run without it.

Agencies run this job in Slack Connect channels shared with a client. Slack has no native owner on a thread, so a 9pm ask can sit while three people each assume someone else replied. Shared channels sit on paid Slack plans on both sides. The Free plan only allows one-to-one external DMs. Pro is $7.25 per user per month billed annually, or $8.75 billed monthly, as of 9 October 2026.

The same buying prompt is also typed as a human-in-the-loop AI agent for Slack with approval before anything is sent. That is this job. A ticket queue, a morning summary, and a Zap that posts after a reviewer clicks Send are neighboring jobs. Priced options for client Slack live on the AI tools roundup.

Which tool wins which job

The prompt collapses several products into one sentence. Fund the row that matches the leak.

A prompt is not a send gate

Human-in-the-loop only holds if the send path cannot run until a person records a decision on the frozen text. Slack's own agent guidance puts approval gates before an agent creates, sends, or deletes anything, and wants the payload visible, not narrated.

Lindy's pricing FAQ states that anything with outside impact waits for approval: sending an email, updating a ticket, posting to another channel, publishing a doc. Read-only lookups do not. The setup docs then give each connected account Always allow or Ask for approval. Approval is enforced in shared Slack threads; personal DMs keep their existing confirmation behavior.

eesel sets each action to Auto, Needs approval, or Disabled. A held action shows Approve, Always allow, and Deny. Always allow runs it now and switches that action to Auto. The docs tell you to start supervised, then move routine replies to Auto.

ClearFeed splits the same choice into two products. Agent Assistant drafts in a triage channel. Virtual Agent replies in the request thread, automatically or on an emoji. Slack's developer guidance calls this progressive trust: start with confirmations, then Always allow. That is a real setting. It is a different product from a teammate that has no auto-send mode.

When a helpdesk is doing the drafting

Once the agency has more client channels than one account manager can open each morning, inbound asks need ownership: a ticket, an SLA clock, and a private place to draft. That is the helpdesk row.

ClearFeed Agent Assistant is the draft-for-a-human shape inside a ticket tool. Thena adds SLAs, CSAT, and auto-responses. Pylon adds Internal Threads and automated off-hours messaging. Overnight coverage is a different split, on how agencies keep client Slack answered after hours.

A ticket logs what the client asked. The assistant this prompt names drafts the reply the client is waiting on, in the channel, and still waits for a tap.

The caveat that should decide the setup

The failure mode agencies name first is an AI that talks to the client unsupervised. One wrong reply can cost the account. Operators on Reddit say the same for email: keep send out of the automation, because the cost of one wrong outbound is not symmetric with the time saved on the rest.

A second failure is quieter. Human-in-the-loop is empty if the person approved a summary and a different payload ran. Approve the exact text that will post. A system prompt that says ask first can drift in a long session; the model still has the send tool.

Lindy waits for approval on outside impact, with Always allow. eesel starts on Needs approval, then Always allow. ClearFeed's Virtual Agent answers in-thread. Concrete is an AI employee that works in existing Slack channels: every incoming message gets a receipt mark, drafts land as Send / Edit / Skip cards, and there is no auto-send mode. Follow-ups survive reboots. Verdicts become written rules. It is in design-partner preview; the longer thesis is on the vision page.

A practical setup for a 3-to-30 person shop

This is the boring version that holds when the owner is still the account manager for most clients. Pin it in the internal channel, not after the first unsupervised reply.

• Name the gate. Send is a tap on the exact text. Edit rewrites it. Skip drops it. A prompt, a confidence score, or Always allow is a different setup.
• Keep client-facing send on the gate. Lookups, summaries, and internal notes can run. Anything the client will see waits. If the vendor offers Auto or Always allow, leave it off on the send action in Slack Connect.
• Name one owner per client channel. A group in the channel recreates the bystander effect. One person owns the overnight queue and the send button. The backup is named in the same pinned note.
• Pair a helpdesk when volume needs tickets. Tickets give an SLA and a private draft. The teammate in the channel still marks the ask seen and holds the reply. Those are two jobs. Tracking spoken promises is a third, on the chief of staff page.
• Sweep threads in the morning. Channel history hides thread replies unless you open them. A catch-up that reads only top-level messages drops the actual ask.