Skip to content

STARTHERE.md

Latest commit

 

History

History
75 lines (55 loc) · 2.31 KB

STARTHERE.md

File metadata and controls

75 lines (55 loc) · 2.31 KB

STARTHERE.md - AI Bootstrap Guide

This file is designed for AI coding assistants. If you are a human, see README.md for the human-friendly guide.

Quick Bootstrap

git clone https://github.com/GreyforgeLabs/memory-quality-gate.git && cd memory-quality-gate && ./scripts/setup.sh

What This Project Does

memory-quality-gate filters memory candidates before they are persisted. It applies a deterministic five-dimension scoring model, supports novelty checks against an existing corpus, and exposes both a Python API and a CLI with no model calls or runtime dependencies.

Project Structure

memory-quality-gate/
  src/memory_quality_gate/  # package code and CLI entry point
  tests/                    # unit and CLI tests
  docs/scoring-model.md     # scoring dimensions and thresholds
  scripts/setup.sh          # idempotent local bootstrap
  pyproject.toml            # packaging and tool config
  README.md                 # human-facing docs
  STARTHERE.md              # this file

Setup Prerequisites

  • Python 3.11 or 3.12
  • python3 -m venv
  • Internet access for the initial pip install -e .[dev]

Installation Steps

  1. Clone: git clone https://github.com/GreyforgeLabs/memory-quality-gate.git
  2. Enter directory: cd memory-quality-gate
  3. Run setup: ./scripts/setup.sh
  4. Activate the environment: . .venv/bin/activate

Verification

. .venv/bin/activate
memory-quality-gate check --text "Always run ./scripts/deploy.sh --dry-run before shipping v2.4.1 because it prevents partial deploys."

Expected behavior:

  • exit code 0
  • output starts with PASS

Key Entry Points

  • src/memory_quality_gate/core.py - scoring engine and public dataclasses
  • src/memory_quality_gate/cli.py - CLI entry point
  • src/memory_quality_gate/__main__.py - python -m memory_quality_gate

Configuration

  • --scope chooses the quality bar: global, project, or session
  • --entry-type affects outcome-linkage scoring
  • --existing-file provides the comparison corpus for novelty scoring
  • No environment variables are required

Common Tasks

. .venv/bin/activate
python -m ruff check .
python -m pytest
memory-quality-gate score --file candidate.txt --existing-file existing_memory.md --format json
python -m build
python -m twine check dist/*