DocsIntegrationsSlack Integration

Slack Integration

Connect Slack to Nareli to automatically capture work context from your conversations and generate intelligent task and time entry suggestions.

Overview

The Slack integration monitors your selected channels, analyzes conversations with a local AI model, and generates suggestions: tasks to create, tasks to update, and time entries to log.

Runs in the background - processes messages at regular intervals
Entirely read-only - Nareli never posts, reacts, or modifies anything in Slack
Fully local - messages are analyzed by Ollama on your machine, never sent to external servers
Your Slack token is stored in the macOS Keychain, not in a config file

The Slack integration requires the Pro plan (€7/mo) or higher. See Subscription & Billing for plan details.

Connecting Slack

Set up the integration from Settings by entering your Slack user token. Nareli validates the connection and displays your workspace name.

Open Settings > Services tab

Create a Slack app in your workspace (or use an existing one)

Add required scopes: channels:history, channels:read (plus groups:history and groups:read for private channels)

Copy the user token (xoxp-...) and paste it into Nareli

Select channels to monitor in the Listeners tab

Optionally map channels to Nareli projects for better suggestion accuracy

Your Slack token never leaves your machine. It is stored in the macOS Keychain and used exclusively by the local Nareli server.

What the Integration Does

The processing pipeline has four stages that run automatically in the background.

Poll - Fetches new messages from monitored channels via the Slack API

Cache - Stores messages locally for context and offline availability

Analyze - Sends message batches to the local Ollama LLM with your project/task context

Suggest - Generates actionable suggestions for your review

Monitor the pipeline in real time through the system status indicator, which shows queue depth and processing state.

Suggestion Types from Slack

The AI generates four types of suggestions based on your Slack conversations.

Type	When Generated	On Accept
task_create	New work discussed that doesn't match existing tasks	Creates a new task with title, description, and project
task_update	New info about an existing task (status, requirements, deadline)	Updates the existing task with new context
timeentry_create	Extended discussion in a project channel implies time spent	Creates a time entry with start time, duration, and task
info	Relevant context that doesn't require action (decisions, status updates)	Dismissed - informational only

Each suggestion includes a confidence indicator and explanation of why the AI generated it.

Channel-to-Project Mapping

Map Slack channels to Nareli projects so suggestions are automatically assigned to the correct project.

Configured in the Listeners settings for each monitored channel
Mapped channels produce higher-quality, more accurate suggestions
Unmapped channels rely on AI inference from message content
View and manage all mappings on the Learnings page

A common setup: map #proj-website to "Website Redesign" and #proj-api to "API Development". Leave cross-functional channels unmapped.

Privacy & Data Handling

All message data stays on your local machine. Messages are cached in your local database, analyzed by Ollama, and suggestions are stored locally.

Token stored in macOS Keychain - never logged or exported
AI runs locally through Ollama - no data sent to OpenAI, Anthropic, or any cloud provider
Local message cache can be cleared anytime from Settings
Clearing cache does not affect already-accepted suggestions or created tasks
Uninstalling Nareli removes all cached Slack data