idadabhai/directory-map — explained in plain English
Analysis updated 2026-05-18
Generate a categorized map of a multi-project folder so an AI assistant understands it in one read.
Re-run a lightweight script to refresh the project index whenever files change.
Attach the generated summary block to a project's instructions file for future AI sessions.
| idadabhai/directory-map | 0xustaz/streamgate | a-bissell/unleash-lite | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | Python | Python | Python |
| Setup difficulty | easy | hard | hard |
| Complexity | 1/5 | 4/5 | 4/5 |
| Audience | developer | developer | researcher |
Figures from each repo's GitHub metadata at analysis time.
Copy the skill folder into the Claude Code skills directory and run one command.
Directory-map is a Claude Code skill that scans any folder and turns it into a map an AI assistant can read and understand quickly. If you work across many projects with an AI coding assistant, you have probably noticed that every new chat session starts from scratch: the assistant has to rediscover what is finished, what is half built, and what the project even uses. This tool turns that repeated discovery into a one time job, or a job you redo with a single command whenever the project structure changes. Running it produces three outputs. The first is a readable file called DIRECTORY_MAP.md that sorts everything in the folder into seven categories, covering things like live projects, work in progress, early concepts, data pipelines, shared infrastructure, creative assets, and old or archived material. This categorization works regardless of what programming language or type of project you are looking at. The second output is a compact block of structured text appended to that file, meant to be read by an AI in one pass instead of scanning the whole folder each time. The third is a small Python script with no external dependencies that can regenerate the index any time the project changes. The author reports a benchmark showing a full pass rate on their test tasks when the skill is used, compared to a zero percent pass rate without it, across three different types of test projects. Installing it means copying the tool into a specific folder used by Claude Code, then running a short command inside a project. From there, the assistant explores the directory tree, sorts files into the categories, and writes the output files automatically. The generated summary can also be placed at the bottom of a project's own instructions file, so any future AI session gets instant context instead of starting cold. The project is released under the MIT license, meaning it can be used freely, including in commercial projects.
A Claude Code skill that scans a project folder and generates a structured, AI-readable map so any coding assistant gets instant context instead of exploring cold each session.
Mainly Python. The stack also includes Python, Claude Code.
Permissive license, free to use, modify, and include in commercial projects as long as the copyright notice stays.
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.