CommitLog currently supports macOS (Apple Silicon). You need at least 16 GB of Unified Memory to run a local AI model.

Open Settings from the app menu, scroll down, and select Local Models. Click Manage AI Models, then click Download on the recommended model. You can stay on that page while it downloads, or close the dialog and track progress via the status header on the main page.

On first launch (after accepting legal terms), select Use Claude API Instead in the AI Model Setup dialog to walk through the Claude setup wizard — it guides you through creating an Anthropic account, buying credits, generating an API key, and validating it in CommitLog.

You can also set this up at any time: open Settings from the app menu, scroll down, select Claude API, then Setup Claude. Paste your API key and click Save. Your API key is stored and encrypted by your local Keychain — you may need to grant CommitLog permission to access it the first time.

A Job Context represents a role at a specific company. CommitLog uses it to keep your logs and projects organized by job. When you write your first log, CommitLog will prompt you to set your first Job Context (your current role and company).

A second Personal Projects context is created automatically so you can log side projects and open source work separately. You can switch between contexts at any time.

You can create Job Contexts for past roles to record work from previous positions. When switching contexts, choose Select Existing to pick a previously created context rather than creating a new one — this keeps all logs and projects for that role together.

From the Logs page, tap the + button to create a new work log. Write what you did, include tech stack details, achievements, and anything worth remembering. The more detail you add, the better the AI Analysis and resume output will be.

Yes. From the Logs page, tap an existing log to edit or delete it. If you edit a log's content, CommitLog resets its analysis flag automatically so it gets picked up the next time an incremental analysis runs — keeping your projects current.

Use CommitLog's AI features to draft and refine your resume content. The generated output is a starting point, not a final product. Edit freely, iterate often, and make sure the result reflects your own voice.

Go to Settings → Run Initial Analysis. The AI will read all your logs, group them by Job Context, and organize them into projects. You can stay in Settings while it runs or leave and track progress via the status header on the main page. It takes a minimum of 1 distinct log to create a project. After the initial run, you can update your analysis at any time using the Update button.

Phase 1: Project Clustering

The AI reads all your log entries and groups related work into meaningful projects based on recurring themes, technologies, and goals. For example, logs mentioning "OAuth", "login UI", and "session tokens" get grouped into an "Authentication System" project. Each log is assigned to exactly one project, and each project gets a name, description, and a confidence score.

Phase 2: Incremental Updates

When you log new work after the initial analysis, CommitLog runs a lighter incremental pass. It looks at your new logs and decides: does this belong to an existing project, or is it something new? It updates project descriptions as it learns more, and creates new projects only when needed.

What Happens Under the Hood

• Your logs are sent to the AI (local or Claude) in batches
• The AI returns structured data - project names, which logs belong to each, technologies used
• Results are stored on your device and displayed as clickable project cards
• Large log sets are automatically split into chunks and merged

The AI never modifies your original log entries - it only creates the project groupings on top of them. You can always re-run analysis if you're not happy with the results.

Auto-Analysis keeps your projects up to date without you having to trigger it manually.

Toggle it via Settings → Auto-Update. When enabled, CommitLog checks for new logs on app startup on either a daily or weekly basis. You can also configure a minimum number of new logs required before the update runs — so it only triggers when there is meaningful new work to process.

Auto-Analysis uses whichever AI you have active — Claude API if configured, otherwise your local model.

Build Experience Section From Logs uses your analyzed projects to draft a job entry for your resume.

Requirements

• You need at least 3 detected projects tied to the selected job context

How It Works

1. Open the modal in Resume Builder
2. Select 3 to 5 projects from the detected project list for each role in your history
3. CommitLog sends those projects to the AI to generate bullet points
4. The generated bullets are used to build a job section in your resume's Experience section

If you add a job description in Resume Builder (using the Job Description button on the left), CommitLog tailors the first draft to that role.

The generated content is a draft. Review and edit it so the final section matches your voice and the role you're targeting.

CommitLog currently supports the following local models:

CommitLog recommends the best model for your machine based on available Unified Memory. You need at least 16 GB of Unified Memory to use any local model.

CommitLog supports the following Claude models when you connect your own API key:

These are estimates and may change. Always verify current pricing in your Anthropic account before heavy usage.
NOTE FROM DEVELOPER: I recommend using Sonnet over Opus for the way CommitLog uses AI - iterative drafting and refinement.

Yes. CommitLog works offline when you use a local model. If you connect your Claude API key, internet is only required for those generation requests.

Your logs and resume drafts are stored locally on your machine. CommitLog does not include telemetry or analytics collection.

You can read the full policy here: Privacy Policy.

CommitLog supports header, summary, skills, experience, projects, education, and certifications sections. With the exception of the header, sections can be dragged and dropped into any order. Once created, each section is saved to your Section Library — for future resumes you can reuse sections from previous versions instead of starting from scratch.

Click the Job Description button in Resume Builder to paste in the role you are targeting. CommitLog uses your active AI model to extract keywords from the description. Once analyzed:

• An ATS (Applicant Tracking System) compatibility score appears as a percentage alongside your resume, updating as you edit sections
• AI suggestions for Summary, Build Experience Section from Logs, and individual bullet points are tailored to that role

Use the ATS score as a guide — higher scores mean better keyword alignment with the job.

Inside the Summary and Experience bullet editors there is an AI-powered suggestion panel. Click the suggestion button to have your active AI model propose an improved version of your current content. You can:

• Apply the suggestion to replace your draft
• Ask Again to generate a different suggestion
• Dismiss the panel and keep your own wording

For summaries it rewrites the full draft; for bullets it refines the individual bullet point. If a Job Description is set, suggestions factor that in automatically.

Absolutely. Generated content is a draft. You can edit, regenerate, and refine everything before export.

Click Export in the top right of the Resume Builder to save your resume as a formatted .docx Word document. A file save dialog lets you choose where to save it. The exported file uses a professional template and is ready to submit to employers.

Go to Settings → Create Backup to save a timestamped snapshot of your entire database — all logs, resumes, projects, and settings. You can also trigger a backup via Save in the app's dropdown menu.

Yes. CommitLog checks for changes every 2 minutes. If changes are detected, a rolling backup is created automatically. Up to 20 rolling backups are kept before the oldest is deleted.

Go to Settings → View Backups to see all saved backups with their date and file size. Select any backup and click Restore. Before restoring, CommitLog automatically creates a safety backup of your current state — so you can always roll back if needed.

Backups are stored in your local app data folder and are never uploaded anywhere.

Use the Email Support button below and include steps to reproduce, screenshots, and your OS version for the fastest response.