Slack

Chat

The Slack connector is a chat connector that pulls live messages from Slack channels via a Bot Token. It uses db.ingest_chat() to extract claims from message threads.

How it works

Messages are grouped by channel and fetched via the Slack Web API
System messages are filtered out — no archived channels, no join/leave/topic messages
Supports three extraction modes: heuristic, smart, or llm
Rate limited with a 1.2s pause between API calls to stay within Slack rate limits
Each channel becomes a source; messages are threaded and deduplicated

Extraction modes

Mode	Description
`heuristic`	Fast rule-based extraction. No LLM required. Default mode.
`smart`	Enhanced heuristic with pattern matching and entity recognition.
`llm`	Full LLM-powered extraction. Requires a configured curator.

1 Create a Slack App

Go to api.slack.com/apps and create a new app for your workspace.

2 Add Bot Token Scopes

Under OAuth & Permissions, add the following Bot Token Scopes:

channels:read — list public channels
channels:history — read message history from public channels

3 Install to Workspace

Click Install to Workspace and authorize the app. Copy the Bot User OAuth Token (starts with xoxb-).

4 Install the requests package

$ pip install requests

5 Connect

conn = db.connect("slack", token="xoxb-...")
result = conn.run(db)

Parameter	Required	Default	Description
`token`	Yes	—	Slack Bot Token (`xoxb-...`)
`channels`	No	All	List of channel names to fetch
`extraction`	No	`heuristic`	`heuristic`, `smart`, or `llm`
`oldest`	No	`0`	Unix timestamp for oldest messages
`save`	No	`False`	Encrypt and persist token

Basic usage

conn = db.connect("slack", token="xoxb-...")
result = conn.run(db)
print(f"Ingested {result.claims_ingested} claims")

Filtered channels with save

conn = db.connect("slack",
    token="xoxb-...",
    channels=["general", "engineering"],
    extraction="heuristic",
    save=True,
)
result = conn.run(db)

LLM extraction

conn = db.connect("slack",
    token=os.environ["SLACK_BOT_TOKEN"],
    extraction="llm",
    oldest=1700000000,
)
result = conn.run(db)