Tandem-Media/fastapi-alertengine
Add one line to your FastAPI app. Detect latency spikes, error surges, and degraded health. Get WhatsApp or Telegram recovery approvals that require your explic
Self-hosted API monitoring that tells you what to do, not just what happened
tofamba.github.io/fastapi-alertengine/
No GitHub topics on this repo.
- Python95.4%
- HTML2.4%
- JavaScript2.2%
1 Review
12:42 PM
fastapi-alertengine feels like it was built from a real operational pain point, not just as another metrics package. The best part is how little setup it asks for: add instrument(app) to a FastAPI app and you immediately get health and metrics endpoints. That makes the project easy to try, which matters a lot for developer tools.
I also like that it thinks beyond “show me a dashboard.” The focus on p95 latency, error rate, health scoring, Redis-backed history, memory fallback, and human-approved recovery gives it a clearer personality than a generic monitoring wrapper. The WhatsApp/Telegram approval flow is especially distinctive, and the Zimbabwe/mobile-first framing makes the product story memorable.
The biggest thing I’d improve is trust-building. The README makes strong claims about test counts, production use, and adversarial auditing, but those would land better with linked proof: CI run links, a short audit note, screenshots, or examples of real output. I’d also clean up small inconsistencies between GitHub and PyPI and add a CONTRIBUTING or SECURITY file.
Overall, this is a useful and focused project. It already has the bones of something maintainers of smaller FastAPI services would actually reach for when they want incident visibility without adopting a heavy observability stack.