Configuration reference

This page covers all Slack bot configuration options. You can edit these settings at any time from the bot’s edit page in the dashboard.

Bot settings

Bot name

A display name for the bot in the Quant dashboard. This does not affect the bot’s identity in Slack — that is determined by the Slack app (Quant-managed or BYO).

Setup type

Choose between Quant-managed and BYO (Bring Your Own). See Setup types for details.

Allowed channels

Restrict the bot to specific Slack channels. When set, the bot only responds in the listed channels. Leave empty to allow all channels.

This applies to @mentions and keyword triggers in channels. Direct messages are always allowed regardless of channel restrictions.

Session TTL

How long conversation history is retained per thread, from 1 to 90 days (default: 14 days). After a session expires, the next message in that thread starts a fresh conversation with no prior context.

Sessions are stored in DynamoDB with automatic TTL-based expiration.

Keyword monitoring

When enabled, the bot responds to messages containing specific keywords in allowed channels, even without an @mention.

Setting	Description
Enabled	Toggle keyword monitoring on or off (default: off)
Keywords	Comma-separated list of trigger words

Caution

When keyword monitoring is enabled, the bot reads all messages in allowed channels to check for keyword matches. This is surfaced prominently during setup. Ensure your team is aware of this behaviour.

Slash commands

Slash commands let users invoke the bot with a typed command:

• Quant-managed: Fixed commands — /ask and /quant
• BYO: Custom commands — configure any command name in your Slack app settings

Event triggers

Trigger	Default	Notes
@mention in channel	Always on	The primary interaction method
Direct message	Always on	Private conversations with the bot
Slash command	Always on	Configurable command name (BYO only)
Keyword match	Off	Opt-in per bot, requires keyword list

Agent configuration

The bot’s backing agent can be configured directly from the bot editor. These settings are the same as those available in the agent configuration page.

System prompt

Custom instructions for the bot’s agent. If sub-agents are also selected, routing instructions are appended automatically.

Tools

Assign built-in and custom tools to the bot’s agent. Some tools are auto-added based on other selections:

• call_agent — automatically added when sub-agents are selected
• query_vector_db — automatically added when knowledge base collections are selected

Auto-added tools show as checked with an “(auto-enabled)” hint.

Skills

Assign skills to the bot’s agent, grouped by namespace. Use the “select all” toggle to assign all skills in a namespace at once.

Knowledge base

Select vector database collections the bot can search to ground its responses in your data.

Sub-agents

Select agents the bot can delegate to. See Agents and routing for details on how routing works.

Sub-agents are optional. If none are selected, the bot handles all questions directly.

Advanced configuration

Expand the Advanced Configuration section in the bot editor to access these settings.

Model

The AI model used by the bot’s agent. Only models that support tool use are available. Pre-populated with your organisation’s default model.

Temperature

Controls response randomness (0–1). Lower values are more deterministic; higher values are more creative.

Max tokens

Maximum number of tokens in the agent’s response. Leave empty to use the model’s default.

Guardrail preset

Select a guardrail configuration to apply content safety policies to the bot’s responses.

Filter policies

Apply input and output filter policies for content validation and governance.

Long context

Enable extended context windows for models that support it.

Thread and session behaviour

• Every Slack thread maps to one persistent AI session
• New threads create a new session; follow-up messages in the same thread reuse the existing session with full conversation history
• The bot always replies in-thread, never in the top-level channel
• In direct messages, the initial message creates a new thread; subsequent messages in the conversation maintain the same session
• Expired sessions (past the TTL) start a fresh conversation if the user messages again

Managing bots

Viewing bots

Navigate to Slack Bots in the dashboard sidebar. The bot list shows:

• Bot name and status (active/inactive)
• Connected workspace name
• Agent assignment and sub-agent count
• Conversation count and last active time

Editing a bot

Click a bot in the list to open the editor. All settings can be changed at any time. Changes to the agent configuration (system prompt, tools, skills) take effect on the next conversation.

Deleting a bot

Deleting a bot:

• Quant-managed: Revokes the Slack OAuth token (uninstalls the bot from the workspace) and deletes the bot record and its backing agent
• BYO: Deletes the bot record and its backing agent. You should manually remove the Slack app from your workspace

API management

All bot operations are available via the REST API:

Endpoint	Method	Description
/v1/slack/bots	POST	Create a bot
/v1/slack/bots	GET	List all bots (org-scoped)
/v1/slack/bots/{botId}	GET	Get bot details (tokens masked)
/v1/slack/bots/{botId}	PUT	Update bot configuration
/v1/slack/bots/{botId}	DELETE	Delete bot and revoke token
/v1/slack/bots/{botId}/token	POST	Store or update credentials (write-only)
/v1/slack/bots/{botId}/sessions	GET	List thread sessions
/v1/slack/bots/{botId}/stats	GET	Usage statistics (30-day window)

Stats endpoint

The stats endpoint returns aggregate usage data for the last 30 days:

Field	Description
totalMessages	Total messages processed
totalToolCalls	Number of tool executions
totalSubAgentCalls	Number of sub-agent delegations
avgResponseMs	Average response time in milliseconds
activeSessionCount	Currently active thread sessions

Next steps

• Agents and routing — Configure multi-agent delegation
• Use cases — Example scenarios and patterns
• API Reference — Full REST API documentation