TraceLite

Lightweight LLM observability SDK with vector semantic caching using PostgreSQL (pgvector) and Redis.

Features

• Async Telemetry: Logs token usage, cost, latency, and TTFT in background threads.
• Two-Tier Caching: Exact hash caching (Redis) + Semantic vector caching (pgvector).
• Fallback Token Counting: Uses tiktoken if native usage streaming metadata is omitted.

Quick Start

docker-compose up -d
pip install -e .

---

#### `.github/workflows/publish.yml`
```yaml
name: Publish TraceLite to PyPI

on:
  release:
    types: [published]

jobs:
  pypi-publish:
    name: Build and publish TraceLite
    runs-on: ubuntu-latest
    environment:
      name: pypi
      url: https://pypi.org/p/tracelite
    permissions:
      id-token: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v4

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - name: Install build dependencies
        run: |
          python -m pip install --upgrade pip
          python -m pip install build

      - name: Build source distribution and wheel
        run: python -m build

      - name: Publish package distributions to PyPI
        uses: pypa/gh-action-pypi-publish@release/v1