/thinking
Set your personal conversational reasoning-effort preference.
Availability
Section titled “Availability”Public source-edition command; execution remains conditional on installation, server policy and the running service. This reference is built from the repository, not a live Discord registration probe. The builder does not declare explicit context restrictions; handler and policy checks still apply.
Syntax
Section titled “Syntax”/thinking [level:String]Angle brackets indicate required input; square brackets indicate optional input. In Discord, choose the command and fill its named options, rather than typing the brackets.
Parameters
Section titled “Parameters”| Name | Type | Required | Allowed values / bounds | Default when omitted | Description |
|---|---|---|---|---|---|
level |
String | No | No registration choices declared; handler validation applies. | Show the saved preference without changing it. | A reasoning level, or “off” to stop sending one. |
Permissions
Section titled “Permissions”No native default-member permission is declared for this path, and no additional native member permission was identified in its reviewed direct handler.
Fork role access is evaluated using the underlying thinking policy key, even when the slash name is grouped. Native Discord permission checks remain separate from Fork role grants. Sensitive commands fail closed before valid setup; native Administrator bypasses Fork’s role matrix, not Discord’s bot permissions or target hierarchy. See Permission model.
Accepted labels come from ThinkingSupport; see the response for the active provider’s capability. This does not change server scans or background jobs.
Examples
Section titled “Examples”/thinking level:highPrefix equivalent and aliases
f!thinking highDefault prefix: f!; installations can configure another. Handler aliases: dusun, düşün, reasoning, reason. Aliases are prefix names, not additional slash commands.
Grouped prefix routes also delegate to the original handlers, but aliases retain that handler’s argument parsing. Conflicting aliases are rejected as ambiguous. Prefix replies are generally visible in the channel; prefer slash privacy controls for sensitive personal data.
Expected result
Section titled “Expected result”Stores the setting for conversational replies. The configured provider/model may ignore or not support it. Off means no reasoning parameter is sent, not a promise of no reasoning.
Failure cases
Section titled “Failure cases”Invalid effort labels are rejected. Model/provider capability can differ from the saved preference. Check the returned error, current bot status and the relevant server permissions. A timeout or uncertain receipt is not proof that nothing happened; inspect state before repeating a mutation.
Usage limits
Section titled “Usage limits”No fixed command-specific cooldown is declared in this reviewed entry. Discord rate limits, provider budgets, service admission and server policy can still apply; this is not an unlimited-usage promise.
Related commands
Section titled “Related commands”Related guides
Section titled “Related guides”Source edition v1 · Registration, direct handler and shared authorization review are hash-bound in the local documentation validation. This is not a production feature-availability certification.