Managing Minds
This guide covers the day-to-day workflow for managing the active mind stack in a project — inspecting installed minds, activating a subset, changing order, and removing activation.
Install a mind card
Mind cards are installed the same way as regular harness cards:
drwn card add @team/base@^1.0.0
drwn install
drwn install fetches the card from the store and runs drwn write to materialize its contributions. If the card declares hooks, grant trust before writing:
drwn card trust @team/base --hooks
drwn write
Inspect installed minds
drwn mind list
drwn mind list --json
The table shows each installed mind, its version, and whether it is currently active. When no explicit stack has been set, all installed minds are active and a note indicates the default mode:
mind version active
@team/base 1.0.0 yes
@team/frontend 2.1.0 yes
Default: all installed minds are active. Run `drwn mind use` to pin an explicit stack.
Set an explicit active stack
To control which minds are active and in what order:
drwn mind use @team/base
drwn mind use @team/base @team/frontend
Order matters for content composition — earlier cards take precedence. After setting the stack, run drwn write to apply it:
drwn mind use @team/base @team/frontend
drwn write
The active stack is persisted to .agents/drwn/config.json under activeMinds:
{
"version": 1,
"activeMinds": ["@team/base", "@team/frontend"]
}
Change the active order
Pass a new ordered list to replace the current stack:
drwn mind use @team/frontend @team/base
drwn write
Remove all mind activation
drwn mind clear resets to an empty stack without uninstalling the cards:
drwn mind clear
drwn write
After clearing, no mind-card content is projected into downstream tool state. The card bundles remain in the store and can be reactivated with drwn mind use.
Remove a mind card entirely
drwn card remove @team/base
drwn write
This removes the card from card.lock and cleans up its materialized content.
Team-shared mind configuration
Commit the project config to share the active stack across the team:
# config.json is at .agents/drwn/config.json
git add .agents/drwn/config.json
git commit -m "activate team mind cards"
When a teammate runs drwn install after pulling, they get the same stack automatically.
Verify the active configuration
drwn status
drwn doctor
drwn doctor reports hookIssues for any installed mind card that declares hooks but has not been trusted by the current user.
See also
- Minds — the active stack model and default behavior
drwn mind— CLI reference- Authoring Mind Cards — creating a mind card from scratch