Quickstart

This quickstart gets GitPKM running in a few minutes.

Prerequisites

• Git
• Python 3

Optional:

• Conda (only if you want to run scripts in a specific env)

1. Clone Repository

git clone <your-repo-url> pkm
cd pkm

2. Bootstrap Once

bash scripts/setup/bootstrap.sh

This will:

1. install versioned git hooks (.githooks)
2. run the full automation pipeline

3. Run the Pipeline Manually

bash scripts/runtime/pkm_python.sh scripts/automation/run_all.py

In an empty starter repository (no CSV tables yet), generation/indexing steps are skipped by design.

Baseline notes are generated automatically from entity tables once you add CSV data.

Index generation is controlled by schema/automation.json. You can start from schema/automation.example.json and adapt it to your own model.

Dataset names are exact.

Examples:

• python pkm.py new person "Alex" writes to data/person.csv
• python pkm.py new people "Alex" writes to data/people.csv

The tooling does not singularize or pluralize dataset names for you.

4. Confirm Output

Once you add entity data and run the pipeline, baseline notes are created automatically.

If you also configure indexes, generated index pages can exist, for example:

• notes/indexes/all_programs.md
• notes/indexes/program_mentors.md
• notes/indexes/program_mentees.md

Generated note blocks use explicit directives such as:

• <!-- GENERATED START: header -->
• <!-- GENERATED START: list:people -->
• <!-- GENERATED START: table:people -->

5. Optional Runtime Configuration

Edit .env if you want a custom runtime:

PKM_CONDA_ENV=
PKM_PYTHON_BIN=python3

• Keep PKM_CONDA_ENV empty for system Python.
• Set it to a conda env name only if needed.

Next Reading

• CLI
• Data Model and IDs
• Schema Contract
• Automation and Validation