Self-hosted error tracking · open source soon
See every error in one place.
No Sentry bill.
Flask Error Tracker is a free error dashboard for Flask. Embed it with one line, or run it as one small service for all your apps. It also catches errors from JavaScript, Node, Next.js and WordPress sites. Everything is stored in one SQLite file, with nothing else to run.
What it does
Catch it, group it, fix it.
It logs server exceptions, browser errors and anything your other apps send it, then turns each error into a report you can paste into an AI assistant or an issue.
One line in Flask
init_error_tracker(app) registers the dashboard at /error-log and a JSON API,
and logs every unhandled 500 with its stack trace, request URL and method.
Every stack, one dashboard
Run it as a standalone service and post errors from any language to
/api/log-frontend-error. Clients are included for browser JS, Node.js, Next.js and WordPress/PHP.
Repeats are grouped
Identical errors (same category, type, message and project) collapse into one entry, with an occurrence count and first-seen/last-seen times, so a noisy loop shows up as one row.
Debug reports for AI
One click exports a Markdown report with the stack trace, context, occurrences and project details, ready to paste into Claude, ChatGPT or an issue tracker.
Browser error capture
A drop-in script captures uncaught JS errors, unhandled promise rejections, failed
fetch calls and console.error. An optional floating debug button
appears when something breaks.
Your server, your data
It's a single SQLite file on your own machine, with no third-party SaaS, per-event pricing or data leaving your network. Categories and project details are yours to set.
How it works
Two ways to run it
Blueprint inside your app
from flask import Flask from flask_error_tracker import init_error_tracker app = Flask(__name__) init_error_tracker(app) # → /error-log
Each app keeps its own dashboard and database. Logging from your own code uses
get_error_db().log_error(...).
One tracker for many apps
python app.py --port 5100
curl -X POST http://localhost:5100/api/log-frontend-error \
-H 'Content-Type: application/json' \
-d '{"errors":[{"error_type":"TypeError",
"error_message":"boom","project_name":"shop"}]}'
Tag each error with a project name and filter the dashboard by project, category or resolved state.
The dashboard
Stats and category tabs. Database, API, frontend, server, worker, test and content processing are built in, and you can add your own.
Live log stream. New errors appear as they arrive. You can resolve them with notes, delete them or clear the resolved ones.
Project context. Each error carries its project name, domain and local folder, so a report shows exactly where to look.
It doesn't do alerting, release tracking, source maps or performance tracing. It's a focused error log, and it can run alongside a full APM if you have one.
Open source
The code is going public on GitHub.
The public repository and the PyPI package are coming soon. This page will link to them when they're live. Until then, the setup guide shows exactly how the tracker is installed and wired up.