How-to Guides

Recipes for pip install open-fdd, releases, and embedding RuleRunner on pandas.

Full platform (data model, APIs, compose, dashboards): open-fdd-afdd-stack — docs/

Engine

PyPI releases (open-fdd) — tags, trusted publishing, twine check.
The optional openfdd-engine package — openfdd_engine vs open_fdd.engine.
Engine-only deployment and external IoT pipelines — RuleRunner on DataFrames.
Desktop app (under construction) — FastAPI gateway, MCP RAG (REST 8090), start-local, restart MCP after index/bootstrap changes, local ingest, Feather storage, BRICK sync, batched rule runs.
Agent & operator playbook (bridge + MCP) — drivers, cleaning, BRICK modeling, FDD tuning, execution notes for assistants (search_docs).
Toolshed (agent file layout) — toolshed/scratch/ vs toolshed/published/; where Codex writes new code.
Verification — pytest and small smoke checks.
Cloning and porting — portable rules and envs.
Operations (engine) — CI and scheduling notes.
Danger zone — expression safety and sharp edges.
Quick reference — imports and common commands.