Files
chat_backend/llm_be/chat_backend/models.py
T
westfarn 093e5462a5
Deploy Beta / unit-tests (push) Successful in 11s
Unit Tests / test (push) Successful in 11s
Deploy Beta / docker (push) Successful in 30s
Deploy Beta / deploy-beta (push) Successful in 7m19s
Eval harness (#62 P4), status frames, and agentic runs (#63) (#71)
## Summary
- Closes Phase 4 of [#62](#62): `evals/suite.json` (≥40 graded questions), `run_evals` management command, and manually-triggered `.gitea/workflows/run-evals.yml`.
- Emits versioned WS `status` frames during grounded chat (evaluating / searching / reading_sources / refining / writing) for [chat_web_app#96](ai_ml_operations/chat_web_app#96).
- Implements [#63](#63): Redis/Celery optional infra, `AgentRun`/`AgentStep`, tool registry (SSRF-safe `fetch_url`, tenant-scoped docs), LangGraph orchestrator, progress frames, REST `GET/POST /api/agent_runs/…`, gated by `ALLOW_AGENTIC_TASKS` (default off).

## Test plan
- [x] `SKIP_RAG_INIT=1 uv run python manage.py test` for evals, ws frames, agent tools, consumers, grounding
- [ ] Manual: with `ALLOW_AGENTIC_TASKS=false`, chat identical to today
- [ ] Manual: status frames visible in FE with #96 branch
- [ ] Manual (GPU): `python manage.py run_evals --runs 3`
- [ ] Manual: `ALLOW_AGENTIC_TASKS=true` multi-step research prompt creates AgentRun + framesReviewed-on: #71
2026-08-04 04:08:50 -07:00

795 lines
26 KiB
Python

from django.db import models
from django.contrib.auth.models import AbstractUser
from django.utils import timezone
from autoslug import AutoSlugField
from chat_backend.storage import DatabaseStorage
import uuid
# Create your models here.
DB_FILE_STORAGE = DatabaseStorage()
class TimeInfoBase(models.Model):
created = models.DateTimeField(default=timezone.now)
last_modified = models.DateTimeField(default=timezone.now)
class Meta:
abstract = True
def save(self, *args, **kwargs):
if not kwargs.pop("skip_last_modified", False) and not hasattr(
self, "skip_last_modified"
):
self.last_modified = timezone.now()
if kwargs.get("update_fields") is not None:
kwargs["update_fields"] = list(
{*kwargs["update_fields"], "last_modified"}
)
super().save(*args, **kwargs)
class LLMModels(TimeInfoBase):
name = models.CharField(max_length=254, default="")
port = models.IntegerField(help_text="This specifies the port that the LLM runs on")
description = models.CharField(
max_length=512,
default="",
help_text="A description for the LLM. Limit is 512 characters",
)
class Company(TimeInfoBase):
name = models.TextField(max_length=256)
state = models.TextField(max_length=2)
zipcode = models.TextField(max_length=5)
address = models.TextField(max_length=256)
available_llms = models.ForeignKey(
"LLMModels",
on_delete=models.CASCADE,
blank=True,
null=True,
help_text="A list of LLMs that company can use",
)
def __str__(self):
return self.name
class CustomUser(AbstractUser):
company = models.ForeignKey(
Company, on_delete=models.CASCADE, blank=True, null=True
)
is_company_manager = models.BooleanField(
help_text="Allows the edit/add/remove of users for a company", default=False
)
deleted = models.BooleanField(help_text="This is to hid accounts", default=False)
has_signed_tos = models.BooleanField(
default=False, help_text="If the user has signed the TOS"
)
slug = AutoSlugField(populate_from="email")
conversation_order = models.BooleanField(
default=True, help_text="How the conversations should display"
)
def get_set_password_url(self):
from django.conf import settings
base = settings.FRONTEND_BASE_URL.rstrip("/")
return f"{base}/set_password/?slug={self.slug}"
class UserAuthEvent(models.Model):
"""Audit trail for auth / account / subscription actions (user admin)."""
class EventType(models.TextChoices):
PASSWORD_RESET_REQUESTED = (
"password_reset_requested",
"Password reset requested",
)
PASSWORD_SET = ("password_set", "Password set")
INVITE_SENT = ("invite_sent", "Invite sent")
ACCOUNT_DELETED = ("account_deleted", "Account deleted")
SUBSCRIPTION_STARTED = ("subscription_started", "Subscription started")
SUBSCRIPTION_UPDATED = ("subscription_updated", "Subscription updated")
user = models.ForeignKey(
CustomUser,
on_delete=models.CASCADE,
related_name="auth_events",
)
event_type = models.CharField(max_length=64, choices=EventType.choices)
created = models.DateTimeField(default=timezone.now, db_index=True)
detail = models.CharField(max_length=512, blank=True, default="")
ip_address = models.GenericIPAddressField(null=True, blank=True)
class Meta:
ordering = ["-created"]
def __str__(self):
return f"{self.get_event_type_display()} @ {self.created.isoformat()}"
@classmethod
def log(cls, user, event_type, *, detail="", ip_address=None):
return cls.objects.create(
user=user,
event_type=event_type,
detail=detail or "",
ip_address=ip_address,
)
class OutboundEmail(models.Model):
"""Record of emails queued/sent by the app (visible in admin)."""
class Kind(models.TextChoices):
PASSWORD_RESET = "password_reset", "Password reset"
INVITE = "invite", "Invite"
FEEDBACK = "feedback", "Feedback"
class Status(models.TextChoices):
QUEUED = "queued", "Queued"
SENDING = "sending", "Sending"
SENT = "sent", "Sent to SMTP"
FAILED = "failed", "Failed"
id = models.UUIDField(primary_key=True, default=uuid.uuid4, editable=False)
kind = models.CharField(max_length=32, choices=Kind.choices)
status = models.CharField(
max_length=16, choices=Status.choices, default=Status.QUEUED, db_index=True
)
to_email = models.EmailField()
from_email = models.EmailField()
subject = models.CharField(max_length=255)
html_template = models.CharField(max_length=255)
text_template = models.CharField(max_length=255)
context = models.JSONField(default=dict, blank=True)
error_message = models.TextField(blank=True, default="")
user = models.ForeignKey(
CustomUser,
on_delete=models.SET_NULL,
null=True,
blank=True,
related_name="outbound_emails",
)
created = models.DateTimeField(default=timezone.now, db_index=True)
updated = models.DateTimeField(auto_now=True)
sent_at = models.DateTimeField(null=True, blank=True)
class Meta:
ordering = ["-created"]
def __str__(self):
return f"{self.subject}{self.to_email} ({self.status})"
class OAuthIdentity(TimeInfoBase):
"""Linked IdP identity + tokens (SSO now; Drive OAuth reuse later — #11)."""
class Provider(models.TextChoices):
GOOGLE = "google", "Google"
MICROSOFT = "microsoft", "Microsoft"
user = models.ForeignKey(
CustomUser,
on_delete=models.CASCADE,
related_name="oauth_identities",
)
provider = models.CharField(max_length=32, choices=Provider.choices)
subject = models.CharField(
max_length=255,
help_text="OIDC subject (sub) from the identity provider",
)
email = models.EmailField(blank=True, default="")
access_token = models.TextField(blank=True, default="")
refresh_token = models.TextField(blank=True, default="")
token_expires_at = models.DateTimeField(null=True, blank=True)
scopes = models.TextField(blank=True, default="")
raw_profile = models.JSONField(default=dict, blank=True)
class Meta:
constraints = [
models.UniqueConstraint(
fields=["provider", "subject"],
name="uniq_oauth_provider_subject",
),
models.UniqueConstraint(
fields=["provider", "user"],
name="uniq_oauth_provider_user",
),
]
verbose_name_plural = "OAuth identities"
def __str__(self):
return f"{self.provider}:{self.subject}{self.user_id}"
FEEDBACK_CHOICE = (
("SUBMITTED", "Submitted"),
("RESOLVED", "Resolved"),
("DEFFERED", "Deffered"),
("CLOSED", "Closed"),
)
FEEDBACK_CATEGORIES = (
("NOT_DEFINED", "Not defined"),
("BUG", "Bug"),
("ENHANCEMENT", "Enhancement"),
("OTHER", "Other"),
("MAX_CATEGORIES", "Max Categories"),
)
class Feedback(TimeInfoBase):
title = models.TextField(max_length=64, default="")
user = models.ForeignKey(
CustomUser, on_delete=models.CASCADE, blank=True, null=True
)
text = models.TextField(max_length=512)
status = models.CharField(
max_length=24, choices=FEEDBACK_CHOICE, default="SUBMITTED"
)
category = models.CharField(
max_length=24, choices=FEEDBACK_CATEGORIES, default="NOT_DEFINED"
)
def get_user_email(self):
if self.user:
return self.user.email
else:
return ""
MONTH_CHOICES = (
("JANUARY", "January"),
("FEBRUARY", "February"),
("MARCH", "March"),
# ....
("DECEMBER", "December"),
)
month = models.CharField(max_length=9, choices=MONTH_CHOICES, default="JANUARY")
class Announcement(TimeInfoBase):
class Status(models.TextChoices):
default = "DEFAULT", "default"
warning = "WARNING", "warning"
info = "INFO", "info"
danger = "DANGER", "danger"
status = models.CharField(
max_length=7, choices=Status.choices, default=Status.default
)
message = models.TextField(max_length=256)
start_date_time = models.DateTimeField(auto_now=True)
end_date_time = models.DateTimeField(auto_now=True)
class Conversation(TimeInfoBase):
user = models.ForeignKey(
CustomUser, on_delete=models.CASCADE, blank=True, null=True
)
title = models.CharField(
max_length=64, help_text="The title for the conversation", default=""
)
deleted = models.BooleanField(
help_text="This is to hide conversations", default=False
)
def get_user_email(self):
if self.user:
return self.user.email
else:
return ""
def __str__(self):
return self.title
class Prompt(TimeInfoBase):
message = models.CharField(max_length=100 * 1024, help_text="The text for a prompt")
user_created = models.BooleanField(
help_text="True if was created by the user. False if it was generate by the LLM"
)
conversation = models.ForeignKey(
"Conversation", on_delete=models.CASCADE, blank=True, null=True
)
file = models.FileField(
upload_to="prompt_files/",
storage=DB_FILE_STORAGE,
blank=True,
null=True,
help_text="file for the prompt (stored in database)",
)
file_type = models.CharField(
max_length=16,
blank=True,
null=True,
help_text="file type of the file for the prompt",
)
citations = models.JSONField(
default=list,
blank=True,
help_text=(
"Structured source citations for grounded answers (#62). "
"List of {index, title, url, published_at}."
),
)
def get_conversation_title(self):
if self.conversation:
return self.conversation.title
else:
return ""
def file_exists(self):
return self.file != None and self.file.storage.exists(self.file.name)
class PromptFeedback(TimeInfoBase):
"""Per-message thumbs rating for an assistant Prompt (chat_backend#67).
Distinct from app-wide ``Feedback`` (product bugs). Joinable to
``PromptMetric`` via ``prompt_id`` for per-model accuracy slices.
"""
class Rating(models.TextChoices):
UP = "up", "Up"
DOWN = "down", "Down"
class Reason(models.TextChoices):
INCORRECT = "incorrect", "Incorrect"
OUT_OF_DATE = "out_of_date", "Out of date"
DIDNT_FOLLOW_INSTRUCTIONS = (
"didnt_follow_instructions",
"Didn't follow instructions",
)
UNSAFE = "unsafe", "Unsafe"
OTHER = "other", "Other"
prompt = models.ForeignKey(
Prompt,
on_delete=models.CASCADE,
related_name="prompt_feedbacks",
help_text="Assistant prompt being rated",
)
user = models.ForeignKey(
CustomUser,
on_delete=models.CASCADE,
related_name="prompt_feedbacks",
)
rating = models.CharField(max_length=8, choices=Rating.choices)
reason = models.CharField(
max_length=64,
choices=Reason.choices,
blank=True,
null=True,
)
comment = models.TextField(max_length=1024, blank=True, null=True)
class Meta:
constraints = [
models.UniqueConstraint(
fields=("prompt", "user"),
name="uniq_prompt_feedback_prompt_user",
)
]
def __str__(self):
return f"PromptFeedback(prompt={self.prompt_id}, user={self.user_id}, {self.rating})"
class PromptMetric(TimeInfoBase):
PROMPT_METRIC_CHOICES = (
("CREATED", "Created"),
("SUBMITTED", "Submitted"),
("PROCESSED", "Processed"),
("FINISHED", "Finished"),
("MAX_PROMPT_METRIC_CHOICES", "Max Prompt Metric Choices"),
)
prompt_id = models.IntegerField(help_text="The id of the prompt this matches to")
conversation_id = models.IntegerField(
help_text="The id of the conversation this matches to"
)
event = models.CharField(
max_length=26, choices=PROMPT_METRIC_CHOICES, default="CREATED"
)
model_name = models.CharField(max_length=215, help_text="The name of the model")
start_time = models.DateTimeField()
end_time = models.DateTimeField(blank=True, null=True)
prompt_length = models.IntegerField(
help_text="How many characters are in the prompt"
)
reponse_length = models.IntegerField(
blank=True, null=True, help_text="How many characters are in the response"
)
tokens_in = models.IntegerField(
blank=True,
null=True,
help_text=(
"Prompt/input tokens reported by the LLM provider usage payload. "
"Null when the provider did not report usage (never estimated)."
),
)
tokens_out = models.IntegerField(
blank=True,
null=True,
help_text=(
"Completion/output tokens reported by the LLM provider usage payload. "
"Null when the provider did not report usage (never estimated)."
),
)
has_file = models.BooleanField(help_text="Is there a file")
file_type = models.CharField(
max_length=16, help_text="The file type, if any", blank=True, null=True
)
def get_duration(self):
if self.start_time and self.end_time:
difference = self.end_time - self.start_time
return difference.seconds
return 0
# Document Models
class DocumentWorkspace(TimeInfoBase):
"""RAG document container: company (business) or user (personal) owned (#46, #55)."""
name = models.CharField(max_length=255)
company = models.ForeignKey(
Company,
on_delete=models.CASCADE,
null=True,
blank=True,
related_name="document_workspaces",
)
user = models.ForeignKey(
"CustomUser",
on_delete=models.CASCADE,
null=True,
blank=True,
related_name="personal_workspaces",
help_text="Set for personal RAG workspaces; null for company workspaces.",
)
class Meta:
constraints = [
models.CheckConstraint(
condition=(
models.Q(company__isnull=False, user__isnull=True)
| models.Q(company__isnull=True, user__isnull=False)
),
name="document_workspace_company_xor_user",
),
models.UniqueConstraint(
fields=["user"],
condition=models.Q(user__isnull=False),
name="uniq_personal_document_workspace_user",
),
]
def __str__(self):
if self.user_id:
return f"DocumentWorkspace(personal user={self.user_id})"
return f"DocumentWorkspace(company={self.company_id})"
class DriveConnection(TimeInfoBase):
"""A linked Google Drive / Microsoft OneDrive-SharePoint account (#47-#52).
``user`` is null for company-only connections set up by a company manager
(kind=company); personal connections always have ``user`` set.
``company`` is required for kind=company. Personal connections may have
``company`` null when the user is not attached to a company (#55).
"""
class Provider(models.TextChoices):
GOOGLE = "google", "Google"
MICROSOFT = "microsoft", "Microsoft"
class Kind(models.TextChoices):
PERSONAL = "personal", "Personal"
COMPANY = "company", "Company"
class SyncStatus(models.TextChoices):
OK = "ok", "Ok"
ERROR = "error", "Error"
PENDING = "pending", "Pending"
NEVER = "never", "Never"
user = models.ForeignKey(
"CustomUser",
on_delete=models.CASCADE,
null=True,
blank=True,
related_name="drive_connections",
help_text="Null for company-only connections owned by manager setup.",
)
company = models.ForeignKey(
Company,
on_delete=models.CASCADE,
null=True,
blank=True,
related_name="drive_connections",
help_text="Required for company connections; optional for personal (#55).",
)
provider = models.CharField(max_length=32, choices=Provider.choices)
kind = models.CharField(
max_length=16, choices=Kind.choices, default=Kind.PERSONAL
)
access_token = models.TextField(blank=True, default="")
refresh_token = models.TextField(blank=True, default="")
token_expires_at = models.DateTimeField(null=True, blank=True)
scopes = models.TextField(blank=True, default="")
external_account_email = models.EmailField(blank=True, default="")
selected_resource_ids = models.JSONField(
default=list,
blank=True,
help_text="Selected folder/drive/site ids to sync (empty = root/default).",
)
selected_resource_labels = models.JSONField(
default=list,
blank=True,
help_text="Human-readable labels matching selected_resource_ids, for the FE.",
)
last_sync_at = models.DateTimeField(null=True, blank=True)
last_sync_status = models.CharField(
max_length=16, choices=SyncStatus.choices, default=SyncStatus.NEVER
)
last_sync_error = models.TextField(blank=True, default="")
# Progress for FE progress bar while last_sync_status=pending (#59).
sync_total = models.PositiveIntegerField(
default=0,
help_text="Remote files discovered for the current/last sync run.",
)
sync_processed = models.PositiveIntegerField(
default=0,
help_text="Files finished in the current/last sync (includes skips).",
)
sync_added = models.PositiveIntegerField(default=0)
sync_updated = models.PositiveIntegerField(default=0)
sync_failed = models.PositiveIntegerField(default=0)
is_active = models.BooleanField(default=True)
class Meta:
constraints = [
models.CheckConstraint(
condition=(
models.Q(kind="personal", user__isnull=False)
| models.Q(
kind="company",
company__isnull=False,
user__isnull=True,
)
),
name="drive_connection_kind_owner_consistency",
),
models.UniqueConstraint(
fields=["user", "provider"],
condition=models.Q(kind="personal"),
name="uniq_personal_drive_connection_user_provider",
),
models.UniqueConstraint(
fields=["company", "provider"],
condition=models.Q(kind="company"),
name="uniq_company_drive_connection_company_provider",
),
]
def __str__(self):
return (
f"DriveConnection({self.provider}/{self.kind}) "
f"company={self.company_id} user={self.user_id}"
)
class Document(TimeInfoBase):
class Source(models.TextChoices):
UPLOAD = "upload", "Upload"
GOOGLE_DRIVE = "google_drive", "Google Drive"
ONEDRIVE = "onedrive", "OneDrive"
SHAREPOINT = "sharepoint", "SharePoint"
GOOGLE_SHARED_DRIVE = "google_shared_drive", "Google Shared Drive"
workspace = models.ForeignKey(DocumentWorkspace, on_delete=models.CASCADE)
file = models.FileField(
upload_to="documents/",
storage=DB_FILE_STORAGE,
help_text="uploaded document bytes (stored in database)",
)
uploaded_at = models.DateTimeField(auto_now_add=True)
processed = models.BooleanField(default=False)
active = models.BooleanField(default=False)
source = models.CharField(
max_length=32, choices=Source.choices, default=Source.UPLOAD
)
remote_file_id = models.CharField(max_length=255, blank=True, default="")
remote_etag = models.CharField(max_length=255, blank=True, default="")
remote_name = models.CharField(max_length=512, blank=True, default="")
drive_connection = models.ForeignKey(
DriveConnection,
on_delete=models.SET_NULL,
null=True,
blank=True,
related_name="documents",
)
sync_error = models.TextField(blank=True, default="")
class Meta:
indexes = [
models.Index(fields=["drive_connection", "remote_file_id"]),
]
class AgentRun(TimeInfoBase):
"""A long-running, multi-step agentic task turn (#63).
``user``/``company`` mirror the tenant scope of the triggering chat turn
(never trust a bare ``conversation_id`` — see ``chat_tenant_scope``).
Progress is broadcast on the Redis channel-layer group
:meth:`channel_group_name` so a reconnecting client can resubscribe.
"""
class Status(models.TextChoices):
PENDING = "pending", "Pending"
PLANNING = "planning", "Planning"
RUNNING = "running", "Running"
COMPLETED = "completed", "Completed"
FAILED = "failed", "Failed"
CANCELLED = "cancelled", "Cancelled"
user = models.ForeignKey(
CustomUser,
on_delete=models.CASCADE,
related_name="agent_runs",
)
company = models.ForeignKey(
Company,
on_delete=models.SET_NULL,
null=True,
blank=True,
related_name="agent_runs",
)
conversation = models.ForeignKey(
"Conversation",
on_delete=models.SET_NULL,
null=True,
blank=True,
related_name="agent_runs",
)
prompt = models.ForeignKey(
"Prompt",
on_delete=models.SET_NULL,
null=True,
blank=True,
related_name="agent_runs",
help_text="The user Prompt that triggered this run, if any.",
)
goal = models.TextField(help_text="Natural-language user request/goal.")
title = models.CharField(
max_length=255,
blank=True,
default="",
help_text="Short human-readable title (from the plan, or the goal).",
)
status = models.CharField(
max_length=16, choices=Status.choices, default=Status.PENDING, db_index=True
)
plan = models.JSONField(
default=list,
blank=True,
help_text="Ordered list of {step_id, title, tool} planner steps.",
)
result = models.TextField(
blank=True, default="", help_text="Final synthesised answer."
)
error = models.TextField(blank=True, default="")
model_orchestrator = models.CharField(max_length=215, blank=True, default="")
model_subagent = models.CharField(max_length=215, blank=True, default="")
max_plan_steps = models.PositiveIntegerField(default=8)
max_iterations = models.PositiveIntegerField(default=12)
wall_clock_seconds = models.PositiveIntegerField(default=600)
tool_call_count = models.PositiveIntegerField(default=0)
iteration_count = models.PositiveIntegerField(default=0)
cancel_requested = models.BooleanField(
default=False,
help_text="Set by the cancel endpoint/frame; worker loop polls this.",
)
started_at = models.DateTimeField(null=True, blank=True)
completed_at = models.DateTimeField(null=True, blank=True)
class Meta:
ordering = ["-created"]
def __str__(self) -> str:
return f"AgentRun({self.pk}, user={self.user_id}, {self.status})"
def channel_group_name(self) -> str:
"""Redis channel-layer group so reconnecting clients get updates (#63)."""
return f"agent_run_{self.pk}"
@property
def is_terminal(self) -> bool:
return self.status in {
self.Status.COMPLETED,
self.Status.FAILED,
self.Status.CANCELLED,
}
def mark_cancelled(self) -> None:
self.cancel_requested = True
self.status = self.Status.CANCELLED
self.completed_at = self.completed_at or timezone.now()
self.save(
update_fields=[
"cancel_requested",
"status",
"completed_at",
"last_modified",
]
)
class AgentStep(TimeInfoBase):
"""A single planner step (optionally decomposed into sub-agent steps)."""
class Status(models.TextChoices):
PENDING = "pending", "Pending"
RUNNING = "running", "Running"
COMPLETED = "completed", "Completed"
FAILED = "failed", "Failed"
SKIPPED = "skipped", "Skipped"
CANCELLED = "cancelled", "Cancelled"
run = models.ForeignKey(
AgentRun,
on_delete=models.CASCADE,
related_name="steps",
)
parent_step = models.ForeignKey(
"self",
on_delete=models.CASCADE,
null=True,
blank=True,
related_name="sub_steps",
help_text="Set when this step was produced by a sub-agent (#63).",
)
index = models.PositiveIntegerField(default=0, help_text="Order within the plan.")
title = models.CharField(max_length=255, blank=True, default="")
status = models.CharField(
max_length=16, choices=Status.choices, default=Status.PENDING, db_index=True
)
is_subagent = models.BooleanField(default=False)
tool_name = models.CharField(max_length=64, blank=True, default="")
tool_input = models.JSONField(default=dict, blank=True)
tool_output = models.TextField(
blank=True,
default="",
help_text="Truncated to AGENT_TOOL_OUTPUT_MAX_CHARS.",
)
error = models.TextField(blank=True, default="")
started_at = models.DateTimeField(null=True, blank=True)
completed_at = models.DateTimeField(null=True, blank=True)
class Meta:
ordering = ["index", "created"]
def __str__(self) -> str:
return f"AgentStep(run={self.run_id}, index={self.index}, {self.status})"
class StoredFile(TimeInfoBase):
"""Blob store for DatabaseStorage — prompt attachments and documents."""
name = models.CharField(max_length=512, unique=True, db_index=True)
content = models.BinaryField()
size = models.PositiveBigIntegerField(default=0)
content_type = models.CharField(max_length=255, blank=True, default="")
def __str__(self):
return self.name