Jev for Slack: AI-Powered Message Classification and Action Automation

Overview

Jev for Slack is an AI application designed to classify Slack messages and automatically trigger configured actions based on those classifications. Built on TypeSafe's Jev decision model, it allows teams to turn recurring decisions into rules directly within Slack. The tool helps triage messages against specific outcomes to trigger reactions, thread replies, or channel posts. Unlike traditional chatbots that generate conversational text, Jev focuses on structured decision-making. It returns probabilistic answers, such as "Yes" with 96 percent probability, which are then mapped to pre-defined Slack actions. The app does not edit other users' messages. It only adds reactions, replies in threads, or posts to designated channels.

Benefits

Jev for Slack offers several key advantages for teams looking to streamline their communication workflows. First, it automates routine triage and routing tasks. This reduces noise in channels and improves response times without the overhead of complex conversational AI. Second, the tool provides explicit control over data processing. Workspace administrators must enable external model processing in the App Home, and consent is checked before every classification. Third, it ensures safety through variable validation. Unknown Slack variables fail validation, and values used only in action text stay out of the classification request. Finally, the app avoids silent data truncation. If the input exceeds the request budget, the request fails with guidance rather than silently cutting off data.

Use Cases

Jev for Slack is ideal for teams that need to handle high volumes of messages and require consistent decision-making. Administrators can create rules via a three-step wizard in the Slack App Home. They choose the conversation by selecting accessible channels and defining triggers like new messages or edits. Next, they describe the distinction by writing a question and defining outcomes. Users can insert explicit Slack variables, such as message text or user ID, into the prompt to provide context. Finally, they close the loop by attaching specific actions to each outcome. These actions can include adding an emoji reaction, posting a configured response in the message thread, or sending a message to a destination channel. This setup is perfect for automating support ticket routing, content moderation, or project status updates.

Pricing

Jev for Slack offers a tiered pricing model based on active human members. The cost is $5 USD per active human member per month, excluding VAT. There is no workspace minimum. The plan includes a two-week free trial. The first week requires no payment method and includes limits of 500 requests and 5 automations. The second week requires a payment method but lifts the request limits. The included features cover model processing costs, 1,800 Jev requests per member per month pooled across the workspace, and 3 automations per billed member. Active members only count as guests, bots, and deactivated accounts are free. The underlying model economics are also efficient, with input costs of $0.042 per million input tokens and free output tokens.

Vibes

Public reception highlights the tool's practicality for enterprise workflow automation. Users appreciate its focus on explicit variables and controlled actions. However, reviews note the probabilistic nature of the AI. Classifications can be wrong, so administrators are advised to use dry tests and keep humans in the loop for critical decisions. The tool is not suitable for high-stakes applications where decision errors carry significant risk without human oversight. Users also note that the app does not backfill old messages, so rules only apply to future communications. Overall, the community sees it as a streamlined solution for reducing noise and improving efficiency when used with appropriate safety checks.

Additional Information

Jev for Slack runs on Cloudflare Workers with D1 storage. The AI processing occurs outside Slack via OpenRouter connecting to TypeSafe's Jev model. Slack methods use a slack_web prefix for custom integrations. The app includes 14-day retention for activity logs, pending work, and action receipts. Recovery features allow captured, eligible events to wait for service recovery within the original deadline. Logging excludes full message bodies and expanded action text to preserve privacy. The tool is vulnerable to prompt injection attacks, so users should be aware of this limitation when designing their rules.