Lightweight LLM observability SDK with privacy-first defaults
Project description
Cecil Python SDK
cecil-sdk is a privacy-first Python SDK for LLM cost visibility and cache optimization insights.
Install
pip install cecil-sdk
Quickstart
import cecil
cecil.patch()
Default behavior is local-only. No telemetry is sent unless explicitly enabled.
Why Cecil
- Privacy-first defaults (hashed metadata, no raw prompt export by default)
- Fail-open instrumentation (SDK failures do not break provider calls)
- Actionable cost and cache opportunity analytics
- Lightweight integration (
import cecil; cecil.patch())
Development
pip install -e ".[dev]"
make lint
make typecheck
make test
make build
python -m twine check dist/*
python scripts/smoke_check_wheel.py
Documentation
docs/quickstart.mddocs/telemetry-opt-in.mddocs/privacy-contract.mddocs/recommendations.mddocs/release-checklist.md
License
MIT. See LICENSE.
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
cecil_sdk-0.1.0.tar.gz
(14.0 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
cecil_sdk-0.1.0-py3-none-any.whl
(17.9 kB
view details)
File details
Details for the file cecil_sdk-0.1.0.tar.gz.
File metadata
- Download URL: cecil_sdk-0.1.0.tar.gz
- Upload date:
- Size: 14.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ea4df070fcaf6f3fde1d9419464801a5391185cee370f4e551527f53a945c0ff
|
|
| MD5 |
471c90ae925c4cc6997739cfee6be881
|
|
| BLAKE2b-256 |
9675d28289d9a82210ad24fc32a6bc40a0a7f006bcb2b1aa27760712f4228bc2
|
File details
Details for the file cecil_sdk-0.1.0-py3-none-any.whl.
File metadata
- Download URL: cecil_sdk-0.1.0-py3-none-any.whl
- Upload date:
- Size: 17.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
e1361b0b71037f85328a36c24617eca712fd2fde3ea12b5676f2f66a847b6de3
|
|
| MD5 |
b613193a2923db1dab5a3ef84efd6ab2
|
|
| BLAKE2b-256 |
129947c28e555623291467e1df41b321c90b41c7934fd20c8f4702f1c096c201
|
