L Local-first by design

Local-first observability for AI-assisted engineering workflows

Understand every step your AI workflows take. Trace sessions, score verification, surface risks, and keep everything on your machine.

OS Open Source
F FastAPI
DB SQLite cache
P Local-first privacy
AI Workflow Observatory
OverviewSessionsTraceRisksReports
Engineering loopworkflow phases

Everything you need to understand your workflows

WF
Workflow phase tracking
Automatically detects phases from context gathering to handoff across every session.
V
Verification quality scoring
Scores verification steps to highlight quality and reliability.
R
Risk flags
Surfaces risky patterns and failures so you can act with confidence.
S
Session trace inspection
Drill into any session and explore detailed logs, events, and artifacts.
EX
Markdown and JSON export
Export insights and traces for sharing, reporting, or offline analysis.
LK
Local-first privacy
All data stays on your machine. No cloud telemetry. Total control.

How it works

1
Scan logs
The Observatory parses your workflow logs, events, and artifacts locally.
-->
2
Classify workflow
It detects phases, scores verification, and flags risks automatically.
-->
3
Review evidence
Explore sessions, traces, and metrics to improve with confidence.
AI Workflow ObservatoryOverview
Explore the Observatory
Open the local demo and start understanding your AI workflows today.