Documentation Index
Fetch the complete documentation index at: https://docs.wednesdayai.dev/llms.txt
Use this file to discover all available pages before exploring further.
Slash commands
Commands start with / and are sent as standalone messages (or embedded in a message for inline shortcuts). They control the current session — switching models, adjusting reasoning depth, managing context, and more.
Commands only work if you are on the allowlist. If a command has no effect, your administrator may need to grant you access.
Most useful commands
Session management
| Command | Effect |
|---|
/reset | Start a fresh conversation (new session) |
/new | Same as /reset |
/new <model> | Start fresh with a specific model |
/compact [instructions] | Summarise context and continue in a smaller window |
/stop | Stop the current running response |
/restart | Restart the agent |
Model and reasoning
| Command | Effect |
|---|
/model <name> | Switch AI model for this session |
/think <level> | Set reasoning depth: off, low, medium, high, xhigh |
/verbose on/off | Show or hide detailed tool output and reasoning |
/think levels and aliases:
/think off — no extended reasoning/think medium — balanced (alias: /t medium)/think high — deeper reasoning, slower responses/think xhigh — maximum reasoning
| Command | Effect |
|---|
/status | Gateway status, active model, provider quota |
/help | List available commands |
/whoami | Show your sender ID |
/context | Show how context is being used |
/usage tokens | Show token count on each reply |
/usage cost | Summarise cost from session logs |
| Command | Effect |
|---|
/exec host=<target> security=<level> | Configure tool execution policy |
! <command> | Run a shell command on the gateway host (if enabled) |
/bash <command> | Same as ! <command> |
/approve <id> allow-once | Approve a pending tool execution |
Voice and output
| Command | Effect |
|---|
/tts always | Always send audio responses |
/tts off | Disable audio |
/tts tagged | Audio only for responses tagged with [voice] |
Sub-agents and sessions
| Command | Effect |
|---|
/subagents list | Show running sub-agents |
/kill <id> | Stop a specific sub-agent |
/steer <id> <message> | Send a steering message to a running sub-agent |
Skills
| Command | Effect |
|---|
/skill <name> [input] | Run a named skill |
Directives in messages
Directives are commands embedded in a normal message. They are stripped before the AI sees the message and adjust the session settings:
/think high /verbose on
Analyse this Python script and suggest improvements.
Inline shortcuts
These commands work when embedded in a message and are stripped before the AI sees the rest:
/help — trigger a help reply; rest of message continues normally/status — trigger a status reply/whoami — show your sender ID/commands — list available commands
Example: Hey /status and what's the weather in Sydney today?
Notes
- Commands and directives only work for authorized senders — users on the allowlist or who have paired with the gateway.
- Unauthorized users see
/... tokens treated as plain text.
- Most commands also work as native commands on Discord and Telegram (tap the
/ button in the chat UI).
/verbose and /reasoning can expose internal tool output — keep them off in shared group chats.
