8b84378fe1
Elevate MiroFish/CrowdSight from single-container dev to a SaaS foundation: - Local memory backend (Zep-compatible): memory services/models, local graph builder + updater, AgentActivity seam, import-boundary isolation; Zep stays default, local is opt-in behind MEMORY_BACKEND. Semantic parity not yet proven. - Durable product persistence: projects/simulations/reports schema (migration 0007) + tenant/owner-scoped ProductRepository + dual-write + scoped_project read-first + ArtifactStore abstraction; durable JobQueue + worker.py. - SaaS hardening: durable RateLimiter (wired to login), UsageService (LLM accounting), redacted AuditService, idempotency, CORS allowlist, safe API errors, single-use PasswordResetService + endpoints (covers invite-pending). - Exactly 3 roles (super_admin/admin/user) with tenant authz policy. - Admin UI: GET/POST/PATCH /api/admin/users + GET/PUT /api/admin/settings (super-admin only, encrypted/masked); AdminView.vue + SettingsView.vue with admin/super-admin route guards, th/en i18n. - Production deploy topology: multi-stage Dockerfile (frontend build + gunicorn wsgi + nginx SPA-proxy + supervisord worker), backend/wsgi.py, gunicorn dep. Backend 197 passed; frontend 10 tests + build green. ruff unavailable (gap). No commit of credentials; secrets handled via env/.env.example. Deferred: Zep semantic A/B parity, object storage cutover, mobile QA, EasyPanel container build of deploy topology.
93 lines
2.8 KiB
Python
93 lines
2.8 KiB
Python
import json
|
|
|
|
import pytest
|
|
from pydantic import ValidationError
|
|
|
|
from app.services.memory_extraction import (
|
|
MemoryExtractionResult,
|
|
build_extraction_prompt,
|
|
parse_extraction_response,
|
|
)
|
|
|
|
|
|
def test_parse_memory_extraction_result_with_entities_edges_and_evidence():
|
|
raw = json.dumps(
|
|
{
|
|
"entities": [
|
|
{
|
|
"mention": "Alice",
|
|
"canonical_name": "Alice",
|
|
"labels": ["Person"],
|
|
"aliases": ["A"],
|
|
"attributes": {"role": "founder"},
|
|
"summary": "A founder.",
|
|
"confidence": 0.92,
|
|
}
|
|
],
|
|
"edges": [
|
|
{
|
|
"source_entity_ref": "Alice",
|
|
"target_entity_ref": "Bob",
|
|
"relation": "KNOWS",
|
|
"fact": "Alice knows Bob.",
|
|
"attributes": {},
|
|
"valid_at": None,
|
|
"invalid_at": None,
|
|
"expired_at": None,
|
|
"confidence": 0.8,
|
|
"evidence": ["episode-1:0-15"],
|
|
}
|
|
],
|
|
"episode_summary": "A relationship statement.",
|
|
"unresolved_mentions": ["Bob"],
|
|
}
|
|
)
|
|
|
|
result = parse_extraction_response(raw)
|
|
assert isinstance(result, MemoryExtractionResult)
|
|
assert result.entities[0].canonical_name == "Alice"
|
|
assert result.edges[0].relation == "KNOWS"
|
|
assert result.edges[0].evidence == ["episode-1:0-15"]
|
|
|
|
|
|
def test_parse_memory_extraction_result_rejects_unknown_fields_and_bad_confidence():
|
|
payload = {
|
|
"entities": [],
|
|
"edges": [],
|
|
"episode_summary": "summary",
|
|
"unresolved_mentions": [],
|
|
"secret_prompt": "must not be accepted",
|
|
}
|
|
with pytest.raises(ValidationError):
|
|
parse_extraction_response(json.dumps(payload))
|
|
|
|
payload.pop("secret_prompt")
|
|
payload["entities"] = [
|
|
{
|
|
"mention": "Alice",
|
|
"canonical_name": "Alice",
|
|
"labels": [],
|
|
"aliases": [],
|
|
"attributes": {},
|
|
"summary": "",
|
|
"confidence": 1.5,
|
|
}
|
|
]
|
|
with pytest.raises(ValidationError):
|
|
parse_extraction_response(payload)
|
|
|
|
|
|
def test_extraction_prompt_puts_language_instruction_first_and_bounds_episode_size():
|
|
prompt = build_extraction_prompt(
|
|
language="th",
|
|
ontology={"entity_types": ["Person"], "edge_types": ["KNOWS"]},
|
|
episode_text="Alice knows Bob.",
|
|
)
|
|
assert prompt.startswith("IMPORTANT:")
|
|
assert "Return JSON only" in prompt
|
|
assert "entity_refs" in prompt
|
|
assert "Alice knows Bob." in prompt
|
|
|
|
with pytest.raises(ValueError, match="episode_too_large"):
|
|
build_extraction_prompt(language="en", ontology={}, episode_text="x" * 20001)
|