AtoAxisDeveloper time tracking · AI coding analytics
Your coding time. The whole picture.
AtoAxis tracks your coding time by project, language and editor, with AI contributions shown alongside your work. Keep your complete history, connect your existing WakaTime plugin, and see where attribution is known or still missing.
Invite-only Early Access. No payment required to join the list.
- Complete history
- WakaTime compatible
- No source code uploaded
02 — Heartbeats
Every Second Tells a Story.
See how your development time takes shape across projects, tools, and coding sessions.
Each mark in the scene is one duration block. Your editor sends a heartbeat when you type, switch to or save a file, and the server joins nearby heartbeats into blocks. A long gap closes a block instead of padding it, so idle time is never counted as coding time.
51editors through the WakaTime plugin protocol
Works from the editor you already use
Connect your existing WakaTime plugin to track coding time. AI attribution depends on your integration.
- VS Code
- Cursor
- Neovim
- Vim
- Emacs
- Sublime Text
- Zed
- IntelliJ IDEA
- PyCharm
- WebStorm
- PhpStorm
- RubyMine
- CLion
- GoLand
- DataGrip
- DataSpell
- Rider
- RustRover
- AppCode
- Android Studio
- Xcode
- Visual Studio
- Eclipse
- NetBeans
- Notepad++
- TextMate
- Kate
- Kakoune
- Micro
- Nova
- Brackets
- Komodo
- Cloud9
- Geany
- Gedit
- Pulsar
- Jupyter
- Azure Data Studio
- DBeaver
- SQL Server Studio
- Postman
- Unity
- Blender
- Figma
- Sketch
- Canva
- Obsidian
- Processing
- Arduino IDE
- TeXstudio
- Terminal
03 — One day
Every Day Becomes a Ring.
Twenty-four hours around one circle. Each lit arc is a stretch of coding, and the rings behind it are the days before.
- Overview
Your time at a glance
Today, all-time active coding, active days and your first recorded day, with a daily goal tracked against the same timeline.
- A daily or weekly coding-time goal, tracked against the same timeline as every other total.
- Today’s total is built from heartbeat-derived duration blocks; idle gaps close a block instead of padding it.
- A WakaTime-compatible total is available beside AtoAxis’s own count, so an imported history can be compared.
- Activity Timeline
Your rhythm, hour by hour
When the work happened, drilled from a link into a timeline, a heatmap and the file list, filtered by project, language, editor or branch.
- Filters live in the URL: project, language, editor and branch, shareable as a link.
- Activity by hour shows your rhythm; drill from here into the heatmap and the file list.
- Each lane is a day; human, AI and Unattributed blocks keep their own pattern.
From Activity to Understanding.
Turn development activity into meaningful insights about your time, projects, and workflows.
This worked example shows how a developer can read one week without opening several reports. The illustrative account logged 28 hours 10 minutes across seven days, including 5 hours 42 minutes today, and 34% of its measured line changes were attributed to AI. The bars show when activity happened; the tiles summarize duration, attribution and the active project. Those figures are sample data created for this preview, not customer results or a performance benchmark. In a real account, AtoAxis derives duration from nearby editor heartbeats, labels line changes as AI, human or Unattributed, and reports coverage so missing evidence is not silently converted to zero. Use the timeline to find an unusual day, then filter by project, language, editor or branch before reviewing the related file list.
Data source: illustrative AtoAxis sample, not customer data.
04 — Human + AI
Development Has Changed.
Understand the Human + AI Workflow.
Explore how human development and AI-assisted activity contribute to your projects.
Human activity
Changes backed by your own editor activity. Keystrokes and characters added or deleted are counted per project and per day, alongside elapsed time.
AI coding agent activity
Line changes reported by a coding agent’s hook, aggregated per agent and per day. Hooks cover Claude Code, Codex CLI, Gemini CLI, Kilo Code and Kimi Code CLI; Cursor and Copilot need adapters.
Evidence we do not have
Work that cannot be proved either way stays Unattributed instead of being credited to you or to an AI. Missing evidence stays unknown.
Combined development activity. Both systems on one timeline, with the coverage behind every total reported as a percentage rather than guessed.
Attribution is code-change attribution, not a measure of time or quality. It is only as complete as the evidence your tools report.
In the Stack Overflow 2025 Developer Survey, 84% of respondents use or plan to use AI tools in development, yet more distrust the accuracy of AI output (46%) than trust it (33%). That gap is why we attribute AI work with evidence instead of guessing. Source: Stack Overflow Developer Survey 2025: AI
05 — The product
Meet AtoAxis.
The same activity, organised into the views you open every day: an overview of your time, a timeline of your rhythm, your projects, and the Human + AI split.
What AtoAxis tracks
Not just elapsed time: which project and language, how the code was written, and where the data lives.
- Projects
Every project, measured
Time broken down by project, language, editor and branch. Several checkouts can be grouped with project rules and merges.
- Time by project, and the same range broken down by language, editor and branch.
- Regex project rules and manual merges group several checkouts into one project, server-side.
- Human vs AI
Who changed the code
AI-attributed lines compared with human keystrokes and characters, with Unattributed kept separate and coverage shown.
- Line changes split three ways: AI, human and Unattributed.
- Coverage is reported as a percentage; missing evidence is never shown as zero.
- AI line counts are aggregated per agent and per day from each agent’s hook.
Complete history
Every account keeps your full coding history. Nothing is aged out to a recent window, and there is no higher tier to buy for it.
AI attribution in the API
Heartbeats accept an AI agent name and AI line counts. Whatever your tooling reports under that field is aggregated per agent and per day.
AI vs. human breakdown
Lines attributed to an AI agent are counted separately from human keystrokes and characters, so the two can be compared over a date range.
Keystroke-level detail
Keystrokes and characters added or deleted are recorded per project and per day, alongside elapsed time.
Project rules & merges
Regex-based project detection and manual merges, applied server-side so every client stays in sync automatically.
AST-based language detection
Tree-sitter parsing instead of file-extension guessing, so ambiguous extensions like .h are handled correctly.
06 — Begin
See the Bigger Picture of Your Work.
Understand your development activity, discover meaningful patterns, and explore how humans and AI work together.
Early Access is invite-only. Join the list and we will email your API key.
Three steps to your first heartbeat
Roughly five minutes, most of it waiting for the extension to install. Nothing to deploy and nothing to configure on a server.
Generate an API key
Open Settings in the dashboard and choose Generate key. One key covers every editor and AI hook on the account.
wakaclock-cli config set-key YOUR_API_KEYPoint your editor at it
Install the VS Code extension and run AtoAxis: Set API Key. In any editor that already has a WakaTime plugin, change one line in ~/.wakatime.cfg instead.
api_url = https://api.atoaxis.com/api/v1Watch it count
Edit a file. The status bar shows today’s total within seconds, and the dashboard Overview shows the same number the server counted.
wakaclock-cli today
Key takeaways
- AtoAxis records your coding time from editor heartbeats — metadata only, never your source code.
- It implements the WakaTime API, so an editor that already has a WakaTime plugin can point at it, and a WakaTime export can be imported.
- Every account keeps its complete coding history instead of a limited recent window.
- AI, human and Unattributed line changes are separated, with the coverage behind each total reported rather than guessed.
- Early Access: planned pricing is $12/month or $99/year, and checkout is not open yet.
AtoAxis at a glance
- Product:
- Hosted developer time tracking service
- Status:
- Invite-only Early Access — checkout is not open yet
- Planned price:
- $12 per month, or $99 per year
- Protocol:
- WakaTime API, including WakaTime export import
- First-party editor client:
- VS Code
- AI agents with bundled hooks:
- 5 — Claude Code, Codex CLI, Gemini CLI, Kilo Code, Kimi Code CLI
- Languages verified by parsing:
- 9 — Rust, Python, JavaScript, TypeScript, TSX, Go, C, C++, Objective-C
- History retained:
- Complete, with no rolling window
- Stored per heartbeat:
- File metadata and counts — never file contents
- Site languages:
- English and Thai
What is AtoAxis?
AtoAxis is a hosted developer time tracking and AI coding analytics service. It turns editor heartbeats into coding-time reports by project, language, branch and editor, keeps your complete coding history, and supports the WakaTime API. Connected integrations report AI, human and Unattributed changes with attribution coverage. Heartbeats contain activity metadata and counts, not source code. Access is currently invite-only Early Access.
AtoAxis heartbeat payloads contain activity metadata and counts, not source code, prompts or AI responses. Connected AI providers have their own data policies.
How does AtoAxis track your coding time?
Your editor sends a heartbeat when you type, switch to or save a file. Each signal carries metadata such as file path, project, branch, language, editor and operating system — never the contents of the file.
The server groups consecutive heartbeats into duration blocks, and a long gap closes a block rather than padding it, so idle time is not counted as coding time.
Line changes are split three ways: AI, human and Unattributed. Work we cannot prove either way stays Unattributed instead of being credited to you, and the coverage behind every total is reported as a percentage rather than guessed.
What can you do with it?
- See coding hours today, this week or across a custom range
- Break time down by project, language, editor and branch
- Filter to one project or language and drill a timeline, heatmap and file list from a link
- Compare AI-attributed lines with human keystrokes and characters
- Set a daily or weekly coding-time goal and track it against the same timeline
- Group several checkouts with project rules and merges
- Import a WakaTime export into the same dashboard
What does AtoAxis not do yet?
That is the case for AtoAxis. On the other hand, here is what we have not built yet — stated plainly, because we would rather you rule us out now than find the gap after you switch:
- Only the VS Code plugin is first-party; other editors use the WakaTime protocol
- AI hooks cover Claude Code, Codex CLI, Gemini CLI, Kilo Code and Kimi Code CLI; Cursor/Copilot need adapters
- Tokens and costs require provider usage plus a known price; unavailable coverage is not shown as zero
- Nothing outside a connected editor or the AI transcript agent is captured; there is no desktop or browser client
- Analytics are single-account; goals, leaderboards and a presence prototype are available where the server runs PostgreSQL, but full team and organisation dashboards remain in development
- The planned hosted product has no free tier
AtoAxis compared with WakaTime
Only rows where both products could be verified are listed, including cases where WakaTime is ahead. Check the current documentation before deciding.
| Capability | AtoAxis | WakaTime |
|---|---|---|
| Complete history included in the base planWakaTime limits dashboard history to 1 week on Free and 2 weeks on Basic; complete history requires Premium. | Premium | |
| Price for complete historyAtoAxis has one plan with every feature included. WakaTime: Free $0 (1 week), Basic $9/mo (2 weeks), Premium $14/mo (complete). | $12/mo | $14/mo |
| Official editor pluginsOther editors reach AtoAxis through WakaTime-protocol compatibility rather than a plugin we publish. | VS Code | 80+ |
| Automatic AI agent tracking (Claude Code, Codex, Gemini CLI, Kilo Code, Kimi Code CLI)AtoAxis configures bundled hooks for Claude Code, Codex CLI, Gemini CLI, Kilo Code and Kimi Code CLI. Cursor and Copilot still need provider-specific adapters. | ||
| AI token usage and estimated costShown only when provider-reported usage and a known effective price are available; otherwise coverage is unknown. | Conditional | |
| Team and organisation dashboardsFull team and org dashboards are in development. Opt-in leaderboards and a presence prototype exist where the server runs PostgreSQL. | ||
| Free tier |
WakaTime details verified against wakatime.com in August 2026. Check their current documentation before deciding. WakaTime is a trademark of its respective owner and is not affiliated with AtoAxis.
Primary sources: WakaTime pricing, WakaTime official plugins. Next monthly review: September 2026.
One plan. Everything included.
Planned launch pricing: one hosted plan with the complete product. Invite-only Early Access; checkout is not open yet.
- Unlimited projects & heartbeats
- AI vs. human code attribution
- Keystroke & character metrics
- Project rules & merges
- AST-based language detection
- WakaTime data import
- Complete coding history, never trimmed
Planned pricing: $12/month or $99/year. Joining the list records interest; it does not create an account or charge a payment method.
Frequently asked questions
What is AtoAxis?
AtoAxis turns editor heartbeats into coding-time, project, language, branch and editor analytics while keeping source code and conversation content out of the heartbeat.
How does AtoAxis track coding activity?
The editor sends metadata heartbeats and the server groups nearby heartbeats into duration blocks for daily and weekly totals.
Is AtoAxis a WakaTime alternative?
AtoAxis implements the WakaTime API and can import a WakaTime export. Fit depends on the editors and reports you need.
Which editors are supported?
A first-party VS Code extension is available; other editors can connect through an existing WakaTime plugin.
Does it track AI coding tools?
The API accepts AI attribution and bundled hooks support Claude Code, Codex CLI, Gemini CLI, Kilo Code and Kimi Code CLI. Cursor and Copilot need adapters.
How is AtoAxis different from WakaTime?
Two things WakaTime does not do: every account keeps its complete coding history instead of a recent window, and line changes are split three ways — AI, human and Unattributed — with coverage reported as a percentage rather than guessed. AI work is also attributed to a named agent, provider and model, with estimated token cost per project.
Can AtoAxis track time spent by AI agents like Claude Code?
Yes. Bundled hooks for Claude Code, Codex CLI, Gemini CLI, Kilo Code and Kimi Code CLI attribute their edits, and a local agent reads those tools' on-device transcripts so work done with no editor open is still counted. Changes that cannot be proven either way stay Unattributed rather than being assigned to you.
How is coding time calculated?
Close heartbeats form duration blocks and a long gap closes one. Idle gaps are excluded.
Does it store source code?
No. Heartbeats contain file metadata and counts, not source contents, prompts or AI responses.
How much does AtoAxis cost?
Planned launch pricing is $12/month or $99/year. Checkout is not open during Early Access.
Who publishes and reviews this page?
AtoAxis product engineering writes and reviews this page against published product behavior and the primary documentation linked below. The publisher is named here so every capability claim has an accountable owner rather than an anonymous byline.
- Publisher:
- AtoAxis
- Reviewed by:
- AtoAxis product engineering
- Last reviewed:
- Policies and contact:
- About the publisherPrivacyContact
How do we verify these claims?
Every capability described above is checked against the source it comes from rather than a competitor round-up: 9 languages are verified by parsing rather than guessed from a file extension, and 5 AI agents ship with a bundled hook or adapter rather than being claimed without one. The WakaTime comparison in particular was verified against wakatime.com in August 2026, and we re-check it monthly.
- 84% of developers use or plan to use AI tools in their development process, yet more distrust the accuracy of AI output (46%) than trust it (33%). Source: Stack Overflow Developer Survey 2025: AI
- AtoAxis implements the WakaTime API, so an editor with an existing WakaTime plugin can send to it. Source: WakaTime API documentation
- Bundled hooks attribute work done through Claude Code. Source: Claude Code hooks documentation
- Bundled hooks attribute work done through Codex CLI. Source: OpenAI Codex CLI
- Bundled hooks attribute work done through Gemini CLI. Source: Google Gemini CLI
- Language is confirmed by parsing the file, not guessed from its extension. Nine languages have a grammar — Rust, Python, JavaScript, TypeScript, TSX, Go, C, C++ and Objective-C — which is how an ambiguous extension like .h is resolved. Source: Tree-sitter
- Current product boundaries and availability are explained in the AtoAxis public guides. Source: AtoAxis public documentation
Last reviewed:
In conclusion: is AtoAxis right for you?
AtoAxis fits you best if you already code in a WakaTime-compatible editor, you want your complete history rather than a rolling window, and you care how much of the work an AI agent actually did.
It is a weaker fit today if you need a desktop or browser client, if your team needs shared organisation dashboards, or if you work mainly in Cursor or Copilot — those still need adapters.
We plan to charge $12/month or $99/year, but our checkout is not open during Early Access, so joining the list now costs you nothing.
Join the AtoAxis early access list
Be first to try the hosted developer analytics product, receive launch updates, and help shape the next integrations.
Join early accessWhat should you know about AtoAxis?
AtoAxis tracks coding activity from editor heartbeats, with project analytics and explicit AI attribution.
Key takeaways
- Product:
- Hosted developer time tracking and coding analytics.
- Availability:
- Early Access; checkout is not enabled.
- History:
- Complete coding history is included on every account.
- Privacy:
- Heartbeat payloads contain activity metadata, not source code or AI prompts.
How can you use coding activity data?
For example, a developer can compare time spent across two projects during the same week. Project and language breakdowns describe recorded editor activity. Review them alongside completed work and time spent planning, reviewing or helping colleagues. Those activities can matter even when they produce no editor heartbeat.
| Signal | What it supports | What it cannot prove |
|---|---|---|
| Editor heartbeats | Recorded coding activity | All working time |
| Attributed changes | Reported source of changes | Code quality or time saved |
Recommendation
Start with one editor and check that its project and language data match your work. Use the setup guide to connect your client, then review attribution coverage before interpreting AI totals. Coding duration is a record of observed activity; it is not a measure of code quality or individual value.
Trade-offs to weigh
On the other hand, editor heartbeats only see work done in a connected editor. Meetings, design reviews and time spent reading elsewhere stay invisible, so an AtoAxis total is a record of coding activity, not of a whole working day.
“A metric that cannot be established should say so instead of rendering as zero.”
Sources and review
Written and reviewed by Jitwisut Thobut, Founder & developer of AtoAxis. Reviewed . Product claims describe our implementation; examples are illustrative, not customer research.
84% of developers use or plan to use AI tools in development, yet 46% distrust the accuracy of AI output and 33% trust it (Stack Overflow Developer Survey 2025). Source.
- AtoAxis public documentation: documented product behavior and data handling.
- AtoAxis setup documentation: account and editor configuration.
- WakaTime API documentation: editor heartbeats and the compatibility protocol.
- DORA software delivery metrics: delivery outcomes, a separate perspective from editor activity.
- Stack Overflow Developer Survey 2025: AI: how developers use and trust AI tools.