Generate a yearly work summary from your git commits — with optional AI-powered elaboration.
- 🐍 Python-based CLI
- ⚡ Fast and lightweight
- 🔍 Scans all git repos under a directory
- 📊 Builds structured commit data using pandas
- 🧠 Optional AI elaboration (Cursor / Claude / manual)
- 🛠️ First-time interactive setup for easy onboarding
pip install docsmegit clone https://github.com/yourname/docsme-py.git
cd docsme-py
python -m venv venv
source venv/bin/activate
pip install -e .Simply run:
docsmeIf no configuration is found, docsme will guide you through an interactive setup:
Thank you for using docsme, proceed with this interactive setup.
what is your github user_name?
> your-username
want me to find where your projects are stored? [y/n]
> y
I found these directories named "projects":
1. ~/projects
2. ~/code/projects
Select one by number:
> 1
please let me know your preferences for AI elaboration:
a. do not elaborate unless I pass --use-agent
b. elaborate with cursor by default
c. elaborate with claude by default
> c
Your configuration will be saved locally and reused automatically.
docsmeGenerates a report using:
- your configured projects directory
- your GitHub username
- the current year
docsme ~/projectsdocsme ~/projects --since "2024-01-01" --until "2024-12-31"docsme ~/projects --author "your-email@example.com"docsme ~/projects --output report.csv
docsme ~/projects --output report.json
docsme ~/projects --output report.parquetdocsme ~/projects --preview 10docsme --use-agentBehavior depends on your setup preference:
manual→ only runs when--use-agentis passedcursor→ enabled by defaultclaude→ enabled by default
docsme --setupDocsme stores your preferences in:
~/.config/docsme/config.jsonExample:
{
"github_username": "your-username",
"projects_root": "/Users/you/projects",
"ai_elaboration": "claude"
}These values are also exposed as runtime environment variables:
DOCSME_GITHUB_USERNAMEDOCSME_PROJECTS_ROOTDOCSME_AI_ELABORATION
Docsme:
- Recursively finds git repositories under your projects directory
- Runs
git logacross each repo - Builds a pandas DataFrame of commits
- Sorts and filters commit history
- Outputs structured data or feeds it into AI workflows
pip install --upgrade docsmepip uninstall docsmegit clone https://github.com/yourname/docsme-py.git
cd docsme-py
python -m venv venv
source venv/bin/activate
pip install -e .- Requires Python 3.9+
- Requires
gitinstalled and available in PATH - Only repositories with a
.gitfolder will be detected
MIT © lramos0