Pack Profiles

A pack profile is the configuration record that ties a service's AI pack together — linking processors, validators, and documents into a unified, resolved configuration. When the AI pack runs for a service, it reads the active pack profile to know which components to use.

Accessed via client.pack_profiles.

Model properties

Property	Type	Description
`id`	`int`	Profile ID
`service`	`str`	Service reference
`pack_enabled`	`bool`	Whether the Pack system is enabled for this service
`universal_executor_enabled`	`bool`	Whether the Universal Executor layer is active
`embedding_model`	`str`	Embedding model identifier
`chunk_overlap`	`int`	Number of overlapping tokens between document chunks
`max_lines`	`int`	Maximum lines per document chunk
`max_chars`	`int`	Maximum characters per document chunk
`top_k`	`int`	Number of top documents to retrieve per query
`return_all_documents`	`bool`	Return all documents instead of top-k
`cosine_similarity_threshold`	`float`	Minimum similarity score for document retrieval
`query_config`	`dict`	Advanced vector query configuration (`exclude_fields`, `exact_search`)

Methods

`list(**filters)`

Returns a lazy generator of all pack profiles.

Filters:

Filter	Type	Description
`service_id`	`int`	Filter by service
`ordering`	`str`	Sort field — prefix with `-` for descending

for profile in client.pack_profiles.list():
    print(profile.id, profile.pack_enabled, profile.top_k)

`filter(**filters)`

Same as list() but returns all results as a list. Accepts the same filters.

profiles = client.pack_profiles.filter(service_id=45)
print(f"{len(profiles)} profiles for this service")

`get(id)`

Retrieves a single pack profile by ID.

profile = client.pack_profiles.get(5)

print(profile.id)
print(profile.pack_enabled)
print(profile.top_k)
print(profile.chunk_overlap)
print(profile.embedding_model)

`create(data)`

Not supported. Raises NotImplementedError. Use create_for_service() instead.

`create_for_service(service_id, data=None)`

Creates a pack profile for a service.

Parameter	Type	Description
`service_id`	`int`	Service to create the profile for
`data`	`dict`	Optional profile configuration

profile = client.pack_profiles.create_for_service(
    service_id=45,
    data={"pack_enabled": True, "top_k": 5}
)
print(profile.id)

`update(id, data, partial=True)`

Updates a pack profile by its ID. Defaults to PATCH.

Parameter	Type	Description
`id`	`int`	Profile ID
`data`	`dict`	Fields to update
`partial`	`bool`	`True` (default) = PATCH, `False` = PUT

client.pack_profiles.update(5, {"top_k": 10})
client.pack_profiles.update(5, {"pack_enabled": False})

`update_for_service(service_id, data, partial=True)`

Updates the pack profile for a service without needing to look up the profile ID first. Uses the service-scoped endpoint.

Parameter	Type	Description
`service_id`	`int`	Service whose profile to update
`data`	`dict`	Fields to update
`partial`	`bool`	`True` (default) = PATCH, `False` = PUT

client.pack_profiles.update_for_service(service_id=45, data={"pack_enabled": True})
client.pack_profiles.update_for_service(service_id=45, data={"top_k": 10})

`delete(id)`

Not supported. Raises NotImplementedError.

`by_service(service_id)`

Returns all pack profiles for a specific service as a list. Shortcut for filter(service_id=service_id).

profiles = client.pack_profiles.by_service(service_id=45)
for p in profiles:
    print(p.id, p.pack_enabled)

`get_service_config(service_id)`

Returns the resolved configuration that the AI pack will actually use at runtime for a service — with all defaults applied and components resolved. Returns a dict. Use this to confirm what the pack will do before a change instance arrives.

Parameter	Type	Description
`service_id`	`int`	Service to get the resolved config for

config = client.pack_profiles.get_service_config(service_id=45)
print(config)