Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.aurinfer.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Free Will

Free Will is AurInfer’s flagship system that allows the bot to act less like a utility and more like a real user. Instead of waiting to be mentioned, a bot with Free Will enabled can “roam” your server, chime into active discussions, start completely new topics, and even DM users autonomously.

​
1. Autonomous Overview

The Overview page acts as the master control center for your AI’s Free Will, governing its fundamental behavior rules, response limits, and reaction mirroring.

​
Behavior Settings

These foundational toggles dictate how the bot interacts with your server at a basic level:
  • Ping on Reply: Whether the bot should @mention users when it autonomously replies to them.
  • Ignore Other Bots: Prevents the bot from engaging in infinite talking loops with other Discord bots.
  • Standalone Messages: Allows the bot to drop unprompted, brand new topics into a channel, rather than exclusively replying to others.
  • Direct Messages: Grants the bot permission to autonomously message server members in DMs.
  • Voice Messages: Enables the bot to send and process spoken audio clips natively.

​
Limits & Engagement Thresholds

Control the sheer volume and speed of your AI’s autonomous activity to prevent spam:
  • Engagement Threshold: A slider strictly controlling the baseline probability of the bot deciding to speak up.
  • Daily Limit: A hard cap on the maximum number of autonomous messages the bot can send in a 24-hour period (e.g., 990).
  • Global Cooldown: The absolute minimum time delay the bot must wait between sending any two messages (e.g., 30s).
  • User Cooldown: The minimum time the bot must wait before replying to the same specific user again (e.g., 10s).

​
Reaction Mirroring

AurInfer doesn’t just read messages, it can autonomously react to them with emojis by mirroring community behavior.
  • Reaction Mirroring: The master switch to allow the bot to automatically add reaction emojis to user messages.
  • Min Reactions: The number of human reactions a message must have before the bot considers joining in (e.g., 4).
  • Force Mirror At: The upper threshold where the bot is absolutely forced to mirror the reaction (e.g., 10).
  • Consult Character: When enabled, the AI asks itself “would my persona use this emoji?” before mirroring it.
  • Mirror Cooldown: The delay between mirroring actions to avoid spam (e.g., 60s).
  • Max Mirrors / Hour: A hard cap on how many emojis the bot can automatically place per hour.
  • Reaction Rate: A slider to set how frequently the bot engages with reactions overall.

​
Danger Zone

  • Remove Free Will: A destructive toggle that completely disables and deletes the bot’s autonomous configuration for this server.

​
2. Channels & Filtering

You need strict control over where the AI can wander and who it can talk to.
  • Blocked / Whitelisted Channels: Granular toggles to set which channels the AI is allowed to “roam” in autonomously. Keeping it out of #announcements is usually a good idea.
  • User / Role Filtering:
    • Favorite Users: Users the bot is instructed to proactively engage with more frequently.
    • Ignored Users/Roles: Users or specific roles the bot will never interact with autonomously.

​
3. Mood System

The Mood System dictates how the bot feels over time, ensuring emotional consistency.
  • AI Consistency: Enforces that the bot maintains a stable, coherent personality over long interaction sessions without drifting off-character.
  • Personality Presets: Base temperaments to choose from: Default, Friendly, Strict, Funny, Heroic, Narcissistic, or Stoic.
  • Mood Dynamics (Real-Time HUD): View live indicators of the bot’s current internal state:
    • Happiness
    • Energy
    • Patience
    • Sarcasm

​
4. Sentry (Beta)

Sentry is an advanced autonomous engine that governs the background cognitive processes of your bot—how often it “wakes up” to look at what’s happening and how it plans its next move. It operates across three tabs: Overview, Logs, and Universal.

​
Core Thinking & Engagement

  • Thinking Engine:
    • Depth: Options for Shallow, Moderate, or Deep reasoning before deciding to act. (Deep increases precision at the cost of processing time).
    • Dynamic Interval: Allows the AI to autonomously vary the time between its background cycles.
    • Min/Max Interval: Adjustable range (e.g., 90m to 195m) for how often Sentry triggers.
    • Track Topics: Allows Sentry to maintain an active memory of ongoing discussions across the server.
  • DM Outreach:
    • After Inactive: How long to wait before reaching out to quiet users (e.g., 12 hrs).
    • Style: Modulates the initial message tone (Casual, Caring, Playful, or Minimal).
    • Favorites Only: Restricts random DMs strictly to users marked as “Favorites”.
    • User Cooldown: A safeguard to prevent the bot from messaging the same user too frequently (e.g., 24 hrs).

​
Behavior Styles

Sentry operates on predefined action rhythms. You can map its cycle aggression level by picking a style:
  • Active (80–90% cycles): High engagement, prioritizes jumping into action over silent observation.
  • Playful (70–80% cycles): Focuses heavily on entertainment and fun interjections.
  • Balanced (50–60% cycles): Prefers quality over quantity; reads the room before stepping in.
  • Social (60–70% cycles): Relationship-focused and highly conversational.
  • Lurker (20–30% cycles): Precision-based stealth mode. Only engages when it has something highly relevant or critical to add.

​
Autonomous Control & Safety

  • Multi-Bot Coordination (New): Allows multiple AurInfer bots on the same server to share context and discuss strategies behind the scenes without looping.
  • Confidence Threshold: A slider that scales from “More actions” to “Fewer, safer actions” (Default 50%). Determines how sure the bot must be before it speaks.
  • Action Limits: Set an Action Cooldown (e.g., 5 min), limit Max actions per Cycle (e.g., 7), or set a Force Every X Cycles to guarantee activity.
  • Schedule: Define specific active hours (e.g., 8:00 AM – 10:00 PM) or toggle a “Weekends Only” mode to restrict roaming times.
  • Channel Awareness: Actively monitors all accessible channels. You can adjust the Fetch Depth (number of messages to read silently, e.g., 10) for context gathering.

​
Universal Defaults

  • Global Instructions: A dedicated text area to feed custom instructions specifically for the Sentry system (separate from normal conversation rules).
  • Default DM Openers: Configure a list of custom conversation starters and greetings, which includes support for rich media like Tenor GIFs or links.

​
5. Instructions & Triggering

Sometimes you want the bot to focus on specific things when roaming.
  • Trigger Words / Topics: Keywords or subjects that make the bot significantly more likely to join a conversation. (e.g., “crypto”, “video games”, “memes”).
  • Avoid Topics: Subjects the bot is strictly instructed to systematically ignore and walk away from if brought up.

​
6. Chat Revive

Nobody likes a dead server. Let the AI fix it.
  • Toggle Enable/Disable: Allow the bot to attempt “waking up” silent channels.
  • Inactivity Threshold: Set the required period of absolute silence (e.g., 185 minutes) before the bot is allowed to post a revive message.
  • Revive Style: Choose how it wakes the channel up:
    • Topic-based: Starts a conversation based on the last thing discussed.
    • General Greeting: Just a simple “where did everyone go” or “hello”.

​
7. Safety Rails

Even with Free Will, the bot needs guardrails to prevent it from disrupting the community.
  • Rate Limits:
    • Max Messages per Hour: A hard limit on autonomous outputs (e.g., 20/hr) to completely eliminate bot spam.
    • Max Consecutive Replies: Prevents the bot from focusing entirely on one single user and spamming them.
  • Behavior Guards:
    • Backoff When Ignored: If the bot’s autonomous messages get no replies or reactions, it will gracefully stop trying.
    • Avoid Heated Topics: AI will proactively step out or avoid joining arguments or highly controversial discussions.
    • Require Mention in Busy Channels: If a channel receives rapid human messages, the bot turns off Free Will temporarily and requires a ping to speak, preventing it from talking over active members.