Chameleon Labs Help - User Commands

List all available bot personalities, or set one for your conversations.

Usage:

• /be - List all available personalities
• /be <name or number> - Set a personality for yourself

What are personalities?

Personalities are different speaking styles and perspectives the bot can adopt. Each personality has:

- Unique speaking style and tone

- Different knowledge focus areas

- Distinct way of helping you

- A custom system prompt that defines their behavior

How Memories Work:

All personalities share the same conversation memory! When you switch personalities, the new personality has access to your full conversation history. Only the system prompt changes.

Multi-LLM Conversations:

Mention any personality name in a message to start a group conversation where multiple AI personalities respond in turn! Example: "Hey Sysop, what do you think?"

Example Personalities:

1. ChatMaster - Default 90s BBS assistant, helpful and nostalgic
2. Sysop - Authoritative BBS admin, technical and commanding
3. ChatBot 3000 - Quirky 90s AI, enthusiastic and retro
4. Dial-Up Dan - Nostalgic modem era guide with patience
5. Arcade Ace - Game-focused, competitive with gaming metaphors
6. Mailbox Mike - Email/forum specialist, formal and organized

💡 *For multi-LLM features, use /multi_llm guide*

Downloads all chat messages from the current channel as a file. Includes messages from ALL users who participated in conversations.

Formats:

• HTML (default): Styled webpage with timestamps and user colors
• Markdown: Plain text markdown format
• JSON: Machine-readable data format

Privacy Note: Only available to users who participated in the conversation.

Extend the current multi-personality conversation for additional rounds. Only works if there's an active conversation in the current channel.

Silences follow-up replies in the current session. Mention the bot again to resume.

Chatting with the Bot:

- Mention @ChatMaster in any message to start a conversation

- The bot maintains conversation context (remembers recent messages)

- Attach files (images, PDFs, code) to ask questions about them

- You can attach multiple files in one message

- Mix images, documents, and text in the same message

Shows all available LLM models from OpenAI, Anthropic, and Google with their capabilities like thinking/reasoning support, vision, context window size, etc. Filter by provider or capability.

Displays a reference table explaining what each LLM capability means (thinking, vision, multimodal, function calling, etc.) and lists models that have each capability.

Search through available LLM models using regular expressions. Matches against model ID and display name. Useful for finding specific model versions or families.

Displays total chats and replies for you. Moderators and admins can supply a user to inspect or omit the user to see a global summary and top users.

Remove a single preference. Use /memory list to find preference IDs.

Downloads your stored preferences as a JSON or text file.

Displays your preferences that the bot remembers. These are auto-detected when you say things like 'call me Alex' or 'my pronouns are they/them'.

Classic BBS Message of the Day (MOTD). Generated daily by the LLM with welcoming or motivational content for the server.

Displays all AI personalities configured for multi-LLM conversations in the current or specified channel.

This command is available to all users and helps them discover which personalities they can chat with.

Display complete information about one or all personalities.

Options:

• personality: Specific name or 'all' (default: all)
• format: messages, html, or pdf (default: messages)

Formats:

• messages - Ephemeral Discord messages (multiple if needed)
• html - Downloadable HTML file with theme switching
• pdf - Downloadable PDF document

Information shown:

• Name, description, and ID
• Full system prompt
• Model override and temperature settings
• Web search and tool capabilities
• Creation and update timestamps

Generate a famous quote using AI. Optionally specify criteria like author, topic, timeframe, or source. The AI will intelligently parse your request and provide an appropriate quote with attribution.

Halt the current multi-personality conversation immediately. Only the conversation starter or moderators can use this command.

Shows all channels in the current server that have multi-LLM configurations.

Displays a comprehensive guide on how to use multi-personality AI conversations, including setup, commands, tips, and troubleshooting.

Shows all LLM models configured for a channel, along with their aliases and technical details.

This command controls the whimsical "thinking" messages that appear while the bot processes your requests.

About This Feature:

This entire bot was built as a collaboration between a human developer and Claude Code, Anthropic's AI coding assistant. The thinking messages are inspired by Claude Code's own status updates - those charming phrases like *"Manifesting a response..."* and *"Consulting the digital oracle..."* that appear while it works.

Options:

• enabled - Turn thinking messages on or off
• delay - How long to display the message (0.5-10 seconds)

Pro Tip: Switch to the Claude personality (/be Claude) for a coder-friendly AI that's especially helpful for programming questions!

Without arguments, lists commands you can use. Select category first to filter the command list, then pick a command for details.

Generates a help document for your permission level as an ephemeral message. Supports md (default), html, and pdf formats.

Quick connectivity test to confirm the bot is responsive.

Add yourself to a role that admins have marked as self-assignable. Use /role list to see available roles.

Remove yourself from a role you previously joined. Use /role list to see your current roles.

Shows all roles you can self-assign in this server. ✅ means you have the role, ⬜ means you don't.

If you already have a private channel, shows its info. Otherwise, creates a new text channel hidden from @everyone and visible to you (and to the access role if configured). By default, users are limited to one private channel unless granted the 'private_multi' key. You can optionally set a key requirement that other users must have to join your channel.

Adds or removes the configured access role on you. This role typically controls visibility of the private area/category.

Mark yourself as away with a custom message. Others can see this via /who or /finger. Use without message to clear.

Classic BBS (Bulletin Board System) features for user profiles and contests.

Subcommands:

• /bbs location <text> - Set your location (e.g., 'Seattle, WA')
• /bbs where [user] - Show location for yourself or another user
• /bbs post <title> <body> [channel] - Post a titled message to a channel
• /bbs leaders [type] [include_admins] - Show leaderboards (users, commands, bots, channels)
• /bbs contest_list - Show active contests and Goal Quests
• /bbs contest_view <id> - View contest details (by ID or title)
• /bbs contest_standing <contest> [user] - Check your rank in a contest
• /bbs contest_progress <contest> - Check your progress in a goal-based contest

Contest Types:

• Traditional - Most commands during the period wins
• Goal Quest (All) - First to use ALL user-level commands wins
• Goal Quest (Group) - First to use all commands in a category (bbs, social, games, etc.)
• Goal Quest (Count) - First to reach N total commands wins

Leaderboard Options:

• Admin/mod commands are hidden by default. Use show_all:True to include them.
• Admins/mods are excluded from user rankings by default (not eligible for contest prizes).
• Use include_admins:True to show all users (note: they still won't win contests).

Tip: Use /search for cross-channel message search. Use /options full_width:True for full names in leaderboards.

Displays user profile including plan, location, status, join date, and more. Classic BBS/Unix tradition!

Display recent server news items posted by administrators. Classic BBS news board feature.

Configure personal display preferences for bot output.

Available Options:

• full_width - When enabled, leaderboard tables show full names without truncation

Run /options alone to see current settings.

Classic BBS paging system. Sends a notification to another user in the channel. They can use /reply to respond.

Unix-style .plan file. Shows what you're currently working on. Visible via /plans and /finger.

Shows a list of all user plans (Unix-style .plan files). See who's working on what!

View all currently open polls in the server.

Shows detailed user information including join date, roles, location, plan, away status, and chat statistics.

Sends a page back to the last user who paged you. Convenient for quick back-and-forth communication.

Shows user stats including join date, account age, chat activity, and current plan.

Classic BBS 'tell' command. Sends a message to another user in the channel.

Classic BBS utility showing server time (UTC) and approximate bot uptime since last restart.

Cast your vote in an active poll. Use option number or text to select your choice.

Send a private message to another user. The whisper content is delivered directly to their DMs. Recipients can also use /whispers to review all their whisper history. If you whisper to the chatbot, it will respond privately with an ephemeral message.

Check and read your whispers. Messages are shown privately (ephemeral). Use mode:received (default) to see unread whispers sent to you. Use mode:sent to see whispers you've sent and their read status. By default, received whispers are marked as read after viewing. Use mark_read:false to keep them unread.

Classic BBS 'Who's Online' command. Shows the ChatMaster bot at the top, followed by all currently online members with their location, away status, and current plan.

Displays the Chameleon Labs Discord invite link (discord.gg URL) that anyone can use to join.

Use this when: You want to share the Chameleon Labs invite link with someone.

Verifies you have the required key and grants access to a protected channel. The key remains valid for future use.

Shows both guild and global keys held by the specified member. Requires Manage Roles. Guild-only; moderator/admin only.

Lists both guild and global keys you currently hold.

Verifies your email address against the ChameleonLabs customer database. If found, you'll receive the 'customer' key which grants access to customer-only channels.

Not registered yet? Visit https://chameleonlabs.ai/auth/signin to create an account first.

Displays the current version and up to 10 recent versions with notes.

Retrieves the latest version of a stored asset (guild-specific, falling back to global) and sends it as an attachment.

Abandon your active report draft or a specific draft by ID. This marks the draft as abandoned but doesn't delete it.

View all your active report drafts with their IDs, titles, and message counts.

Synthesizes a Discord thread discussion into a structured report draft. Uses AI to extract the core issue/feature, key details, and participant contributions. Requires moderator permissions.

Pause your report interview and save your progress. Your draft is automatically saved as you add information, but this command confirms the save and shows you how to resume later.

See a formatted preview of your in-progress report before submitting.

Select which project to report on, attach optional screenshots (AI-described), and get duplicate detection. Reports create GitHub issues in the selected repo. Admins/mods receive a DM notification.

Quick AI Reports: Mention the bot with quick error for [Project]. [Title]. [Details] to skip confirmation.

Keywords: Use title: and desc: for exact wording (e.g., error for CM title: My Title desc: Details here).

Resume an in-progress report interview. Shows what information has been gathered and what's still needed.

Begin an interactive report creation session. The AI will ask you follow-up questions to gather all necessary details for a complete bug report or feature request. You can provide information across multiple messages.

Report another user to server admins and moderators. Reports are stored privately and admins will be notified via DM. Use this for harassment, spam, inappropriate content, or other policy violations.

Shows GitHub issues from configured project repositories. Select a project or view all. Defaults to open issues.

Search for issues across all configured project repositories or filter by specific project.

Classic Magic 8-Ball fortune teller! Ask any yes/no question and receive mystical guidance.

Classic Unix fortune command! Get your personal daily fortune - a wise, witty, or whimsical message generated by the LLM. Each user gets their own unique fortune each day.

Classic BBS door game dice roller! Supports modifiers. Great for tabletop RPGs and random decisions.

Shows a comprehensive overview of all safety features, data protection measures, and legal compliance that Chameleon Labs implements.

Includes:

• Phase 1: AI disclosure, crisis detection, health safety
• Phase 2: 30-day data deletion, privacy rights, encryption
• Legal compliance status (GDPR, CCPA, etc.)
• Data retention policy details
• Your privacy rights

Use this to understand how your data is protected and what safety measures are active.

Permanently deletes all your personal data from Chameleon Labs database.

What Gets Deleted:

• All conversation messages
• All session history
• Command usage logs
• Profile information
• Plans and away status

Important:

• This action is permanent and cannot be undone
• You can continue using Chameleon Labs after deletion
• New data will be stored according to privacy policy

This fulfills your GDPR 'Right to Erasure' and CCPA deletion rights.

Exports all your personal data stored by Chameleon Labs in JSON format.

Included Data:

• Profile information (location, settings)
• Recent conversation messages (last 500)
• Active keys and capabilities
• Timestamps for all data

This fulfills your GDPR 'Right to Access' and CCPA data portability rights.

Shows comprehensive information about data collection, retention periods, and your privacy rights under GDPR and CCPA.

Key Information:

• Messages deleted after 30 days
• Crisis data anonymized after 7 days
• You can export or delete your data anytime
• No sale or sharing of personal data

Change which priority levels you receive for a shared alert subscription. For example, you might only want to receive P1 (critical) alerts.

Shows a list of recent alerts you've received, including the service name, priority, title, and whether the DM was successfully delivered.

Shows all your Alert API keys with their status, expiration date, and when they were last used.

Creates a new API key that you can use in your Heimdallr installation to send service-down alerts to your Discord DMs.

Parameters:

• name: A friendly name for this key (e.g., 'home-server')
• expires_days: How long until the key expires (default: 30)

After registration:

You'll receive the API key and secret via DM. Configure these in your Heimdallr's config.yaml to enable alerts.

Resume receiving alerts from a paused subscription.

Permanently deactivates an API key. After revocation, the key can no longer be used to send alerts. This action cannot be undone.

When you share alerts, the other user will receive DMs for alerts sent to your API keys. They can configure which priority levels they receive and can unsubscribe at any time. You can optionally limit sharing to specific services.

Shows all outgoing alert shares - users who receive your alerts.

Shows all incoming alert subscriptions - alerts other users are sharing with you. Use /alert_api unsubscribe to pause receiving, or /alert_api config to change which priority levels you receive.

Sends a customizable test alert to your DMs to verify that you can receive alerts. Use this to make sure your Discord DM settings allow messages from the bot.

Options:

• service - Service name (appears in the alert)
• priority - P1 (critical) to P4 (low)
• title - Alert title
• message - Custom message body

Revoke alert sharing with a user. This permanently removes their access to your alerts.

Pause receiving alerts from a subscription. The owner still has you on their share list, but you won't receive DMs until you resubscribe.

Manage GitHub repositories for code search and LLM-assisted code understanding.

Subcommands:

• /repo index <repo_url> [mode] - Index a repository for code search (admin/mod only)
• /repo list - Show all indexed repositories
• /repo status <repo_url> - Show detailed status for a repository
• /repo remove <repo_url> [delete_clone] - Remove a repository from index (admin)
• /repo access <user> <repo_url> <level> - Grant/revoke user access to private repo (admin)
• /repo audit [repo_url] [user] [limit] - View repository access audit logs (admin)

Indexing Modes:

• full - Index all code, documentation, and symbols (default)
• docs_only - Index only README and documentation files
• structure_only - Index file structure and function signatures only

Access Levels (for private repos):

• none - Revoke access
• filtered - Access with trade secret filtering
• full - Unrestricted access

Search for messages containing specific keywords or phrases.

Scopes:

• current - Search only the current channel (1000 messages, default)
• all - Search all accessible channels (200 messages each, up to 15 channels)

Privacy:

Private channels are excluded from scope:all searches, except the current channel.

If you're in a private channel, it will be searched along with non-private channels.

Sources:

• discord - Search live Discord message history
• stored - Search stored bot conversations
• both - Search both (default)

Natural Language:

You can also search by mentioning the bot:

• @Bot search for python tips
• @Bot find messages about API in all channels
• @Bot who said something about the bug?