Give a fresh AI coding session context on past decisions before it starts editing an unfamiliar codebase
Find out which parts of a repository have been rewritten or reverted the most
Check what tests or warnings are currently being silenced in the codebase
See a repository's history told back as a readable story instead of a raw commit list
| promptwheel-ai/logbook | abhagsain/bayut-maps-ratings | abhishek-kumar09/who-is-imposter | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | JavaScript | JavaScript | JavaScript |
| Last pushed | — | — | 2021-04-12 |
| Maintenance | — | — | Dormant |
| Setup difficulty | easy | moderate | easy |
| Complexity | 2/5 | 2/5 | 3/5 |
| Audience | developer | general | general |
Figures from each repo's GitHub metadata at analysis time.
One npx command with no dependencies or accounts needed, large repos just take longer on the first scan.
logbook is a command line tool that reads a project's entire git commit history and turns it into a short briefing document that AI coding assistants can read before they start working on the code. The idea is that fresh AI sessions usually only see the current code, not the story behind it: which parts were rewritten more than once, which approaches were tried and abandoned, or which test was quietly skipped last quarter. Reading every commit yourself would take too much space for an AI's context window, so logbook mines up to the most recent twenty thousand commits and compresses them into a compact record instead. Running one command scans the repository and writes three files. LOGBOOK.md is the short brief meant to be loaded directly into an AI assistant's instructions, covering frequently changed files, approaches that were reverted, and a list of times a test was skipped or a warning silenced. events.jsonl is a more detailed, one line per commit data file that other tools can query. JOURNEY.md tells the repository's history back as a narrative, viewable in color from the command line. The tool works entirely on your own machine. It never sends your source code or git history anywhere, and it only writes its own brief files, plus an optional block added to an existing agent configuration file such as CLAUDE.md so that future AI sessions are told to read the brief first. Scanning a large repository's full history is the slow part, but logbook caches the result so later runs only scan new commits, cutting a forty plus second scan down to a fraction of a second on a repeat run. A companion audit command shows what is still silenced in the current code, and a doctor command checks that the generated files and the agent wiring are still healthy. logbook has zero external dependencies and ships as a single file.
A single file, zero dependency CLI that mines a repo's git history into a compact brief AI coding agents can read before working on the code, so they know about reverted approaches and skipped tests.
Mainly JavaScript. The stack also includes JavaScript, Node.js, npm.
The README does not state a license.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
double-check against the repo, no cap.