Quickstarts

This page consolidates the most important setup and usage patterns across the local BriteCore ecosystem.

Start here depending on the problem:

  • Need live API access -> britecore_sdk
  • Need report/view discovery -> britecore_mcp
  • Need raw export relationship checks -> britecore-csv-loader
  • Need a shared reporting service layer -> PolicyReporting

britecore_mcp

Local install

cd C:\PythonProjects\BriteCore\britecore_mcp
python -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install .

Typical run

cd C:\PythonProjects\BriteCore\britecore_mcp
.\.venv\Scripts\python.exe -m britecore_mcp.server

Best first exploration target

SELECT *
FROM v_logical_catalog;

Use this to locate the logical views and fields that match the business question.

britecore_sdk

Local editable install

cd C:\PythonProjects\BriteCore\britecore_sdk
python -m pip install -e .

Optional extras

python -m pip install -e ".[async-http]"
python -m pip install -e ".[typed-config]"

Typical use pattern

from britecore_sdk.api.api_calls import get_api_client
from britecore_sdk.api.api_calls.v2 import policies

client = get_api_client()
result = policies.retrieve_policy(policy_number="POL001")
print(result)

britecore-csv-loader

Local install

cd C:\PythonProjects\data\britecore-csv-loader
python -m pip install -e .

Relationship inspection

python -m britecore_csv_loader.cli --relationships

Claims extractor example

python -m britecore_csv_loader.part2_claims --reporting-year 2025

Python usage

from britecore_csv_loader import RawCsvInterface

api = RawCsvInterface()
revision_id = api.load_dataset("revisions")[0]["revisionId"]
print(api.revision_summary(revision_id))
print(api.extract_entity_links(revision_id))

PolicyReporting

Typical setup pattern

cd C:\PythonProjects\PolicyReporting
python -m pip install -e .

Query pattern

from policy_reporting import CrossSourcePolicyQuery

query = CrossSourcePolicyQuery.from_env()
result = query.find_by_form("HO 002*")
print(result.summary())

Cross-project starter workflow

For most reporting work, the best flow is:

  1. Identify the business object and grain.
  2. Use britecore_mcp to find logical layers and candidate views.
  3. Verify relationship assumptions from raw CSV exports if needed.
  4. Use britecore_sdk for live API retrieval when required.
  5. Move the final cross-source logic into PolicyReporting when a service boundary is needed.

Practical decision guide

If the ask is "Which table has this data?"

Use britecore_mcp.

If the ask is "Can you pull this from the live API?"

Use britecore_sdk.

If the ask is "I only have raw exported CSVs and need the relationships?"

Use britecore-csv-loader.

If the ask is "I need one service interface across BriteCore and other sources?"

Use PolicyReporting.

Documentation references