Comments
>log in to commentHey everyone, I'm Pavel, the creator of DevGhost. A busy Git history can still leave a simple question unanswered: how much work was actually delivered? That question gets even more interesting when developers start working with AI agents. DevGhost estimates the work behind code changes against a developer-without-AI baseline, then lets you explore contributions and trends. You can run the analysis entirely locally, including the analytics, so your code and results stay on your machine. I'd especially like feedback on the baseline and the charts: what would make these reports more useful when reviewing your own work or discussing delivery with a team? The Ghost Café demo is a quick way to explore them, and the local edition includes a free monthly allowance.
- AI estimates cognitive effort behind code changes
- Shows team output as Ghost% benchmark
- No time tracking, screenshots or keystroke logging
- Provides recommendation zones for conversation
- Free plan available
DevGhost estimates the effort behind Git changes using a mid-level developer benchmark.
- for
- Founders, CTOs, engineering leaders, and developers who need insight into code output.
- pricing
- freemium
Key features
- Git-change effort estimation — Analyzes commit content and estimates work in reference-developer hours.
- Interactive dashboards — Shows contributor breakdowns, timelines, and productivity charts.
- AI adoption comparison — Tracks output changes before and after AI coding assistants.
- Local, cloud, and on-prem modes — Available as SaaS, fully local VS Code and Cursor extensions for Windows and macOS, a Windows desktop beta, and on-premises deployment.
- Free monthly commit quota — 500 new commits analyzed free per month locally; additional capacity through one-time credit packs.
- Recommendation zones — Highlights growth, recognition, context checks, and unblock opportunities.
Use cases
- Compare developers and contractors to a fixed mid-level developer familiar with the codebase, working without AI.
- Assess contractor or vendor contributions using the same metric.
- Evaluate the impact of AI coding assistants on team productivity.
- Identify developers consistently above or below the benchmark for recognition or coaching.
DevGhost pricing
- SaaS (cloud)$02,000 signup credits + 500 each month · Full product available free · Credits purchasable as one-time packs
- Local (Windows/macOS)$0500 new commits free each month · All analysis runs on your machine · No internet or external calls
DevGhost FAQ
Does DevGhost monitor developers with screens or keystrokes?+
No, it only analyzes code changes in Git; no time tracking or surveillance is performed.
Can I use DevGhost without sending code to the cloud?+
Yes, the Local mode runs entirely on your computer with no external calls, keeping source code and results local.
How reliable are the effort estimates?+
Estimates can vary per commit; the tool is intended for aggregated trends and should be combined with context such as role and task complexity.
Is DevGhost suitable for making personnel decisions?+
It should be used as one signal only, not the sole basis for hiring, firing, or compensation decisions.
How does AI usage affect the benchmark?+
The benchmark stays fixed to a mid-level developer without AI, allowing you to compare output before and after AI adoption while accounting for role and task changes.
Edited by the makers, based on devghost.pro · Oct 13, 2026. Details may change; check the official site.
