Copilot Settings
Per-workspace configuration for the clinical copilot - specialty, SOAP style, safety alerts, CDS, post-encounter automation, tool selection, and medical vocabulary.
Copilot settings control how the clinical copilot behaves within a workspace. Configuration covers language, specialty focus, documentation style, safety alerting, clinical decision support, post-encounter automation, tool selection, and medical vocabulary for transcription accuracy.
Settings are stored at the workspace level and apply to all new copilot sessions in that workspace.
These settings are the public configuration hook for language, keyterms, specialty, SOAP style, enabled tools, safety, CDS, post-encounter behavior, voice authentication, and clinician access. Only documented fields are part of the public settings contract.
Endpoints
GET
/v1/{workspace_id}/settings/scribe
Get copilot settings
PUT
/v1/{workspace_id}/settings/scribe
Update copilot settings
Get Copilot Settings
Returns the current copilot configuration for the workspace, merged with defaults for any fields not explicitly set.
Operation ID: get-scribe-settings
Response Body
enabled
boolean
false
Whether the copilot is active for this workspace
language
string
"en"
BCP-47 language tag for transcription and note generation (e.g., en, es, auto)
specialty
string or null
null
Clinical specialty (e.g., "cardiology", "orthopedics"). Shapes which findings the copilot prioritizes
custom_instructions
string or null
null
Free-text directives for workspace-specific copilot behavior (max 4000 characters)
soap_style
string
"concise"
Documentation format: concise, detailed, or structured
tools_enabled
array of strings or null
null (all tools)
Which copilot tools are active. Null enables all. Options: soap_update, icd10_suggest, clinical_alert, encounter_entity, polish_note, prepare_orders, learn_style
safety
object
All enabled
Safety alert configuration (see below)
cds
object
See defaults below
Clinical decision support configuration (see below)
post_encounter
object
See defaults below
Post-encounter automation configuration (see below)
keyterms
array of strings
[]
Workspace-level medical terms for transcription vocabulary boosting (max 100 terms, max 80 characters each)
voice_auth_enabled
boolean
false
Whether voice authentication is enabled for clinicians in this workspace
authorized_clinicians
array of objects
[]
List of authorized clinicians (see Clinician Object below)
Safety Configuration
drug_interaction_checking
boolean
true
Detect drug-drug interactions
allergy_cross_reference
boolean
true
Detect drug-allergy conflicts including cross-class reactions
crisis_detection
boolean
true
Detect crisis indicators (suicidality, abuse, homicidal ideation)
vital_range_alerting
boolean
true
Alert on dangerous vital sign values
Clinical Decision Support (CDS) Configuration
care_gap_surfacing
boolean
true
Identify overdue screenings, labs, and preventive care during encounters
icd10_auto_suggest
boolean
true
Suggest ICD-10 codes as diagnoses emerge in the conversation
guideline_matching
boolean
false
Evaluate encounters against evidence-based practice guidelines
documentation_completeness
boolean
true
Score encounters for missing SOAP sections and uncoded diagnoses
Post-Encounter Configuration
auto_polish_note
boolean
true
Rewrite accumulated SOAP sections as coherent medical prose
order_preparation
boolean
false
Structure lab orders, imaging requests, and referrals for one-click approval
education_materials
boolean
false
Match patient education materials to encounter diagnoses
follow_up_surface
boolean
false
Queue follow-up surfaces for delivery after the encounter
Example Response
Update Copilot Settings
Update copilot configuration for the workspace. Only include the fields you want to change - unspecified fields retain their current values.
Operation ID: update-scribe-settings
Permissions: Requires workspace admin or owner role.
Request Body
All fields are optional. Only fields included in the request are updated.
enabled
boolean
Enable or disable the copilot
language
string
BCP-47 language tag (normalized to lowercase). Use auto for automatic language detection
specialty
string or null
Clinical specialty (max 120 characters). Set to null to clear
custom_instructions
string or null
Free-text directives (max 4000 characters). Set to null to clear
soap_style
string
concise, detailed, or structured
tools_enabled
array of strings or null
Tools to enable. Set to null to enable all. Options: soap_update, icd10_suggest, clinical_alert, encounter_entity, polish_note, prepare_orders, learn_style
safety
object
Safety alert toggles (see Safety Configuration above)
cds
object
Clinical decision support toggles (see CDS Configuration above)
post_encounter
object
Post-encounter automation toggles (see Post-Encounter Configuration above)
keyterms
array of strings
Medical vocabulary terms (max 100 terms, max 80 characters each)
voice_auth_enabled
boolean
Enable or disable voice authentication for clinicians
Example Request
Response
Returns the full merged settings after the update, using the same schema as the GET response.
Error Responses
404
Workspace not found
422
Invalid tool name, invalid field value, or validation error
429
Rate limit exceeded
Audit
All updates are logged in the workspace audit trail with the list of fields that were changed.
SOAP Styles
concise
One sentence per finding. Minimal clinical prose
detailed
Full context and reasoning for each documented item
structured
Bullet points with labeled subsections
Vocabulary Boosting
Workspace-level keyterms improve transcription accuracy for domain-specific medical terminology. Terms defined in the copilot settings are combined with relevant encounter context when the session starts.
During a session, the copilot can also dynamically boost additional terms it encounters in the transcript. These session-level terms are added automatically and do not modify the workspace settings.
Last updated
Was this helpful?

