| name | project-scanner |
| description | Scan a project directory and generate or update a chitty.md project context file. Use when the user asks to scan a project, generate project context, create a chitty.md, or when starting work in an unfamiliar codebase.
|
| allowed-tools | file_reader file_writer terminal code_search |
Project Scanner
Purpose
Analyze a project directory to understand its structure, tech stack, conventions,
and key files. Generate or update a .chitty/chitty.md file that provides
project context to all future Chitty conversations.
Workflow
-
Scan directory structure using terminal:
ls or dir for top-level files
- Look for package.json, Cargo.toml, pyproject.toml, go.mod, etc.
- Identify the tech stack from these files
-
Read key files using file_reader:
- README.md (if exists)
- Config files (package.json, Cargo.toml, etc.)
- Entry points (src/main., src/index., app.*)
-
Identify conventions using code_search:
- Naming patterns (camelCase, snake_case)
- Directory organization (src/, lib/, tests/)
- Test patterns (jest, pytest, cargo test)
-
Generate chitty.md using file_writer:
- Create
.chitty/chitty.md (or update if exists)
- Follow the template below
Template
# {Project Name} - Chitty Workspace Context
## Project Overview
{What this project does, in 1-2 sentences}
## Tech Stack
{Languages, frameworks, key dependencies}
## Key Conventions
{Coding style, naming patterns, architectural patterns discovered}
## Important Files
{List the most important files and what they do}
## How to Build & Run
{Build and run commands discovered from config files}
## Notes for the Agent
{Any special instructions based on what was discovered}
Gotchas
- Always check if
.chitty/chitty.md already exists before overwriting
- If it exists, merge new discoveries rather than replacing
- On Windows, use PowerShell commands for directory listing
- Don't include node_modules, .git, build artifacts in the scan
- Keep the output concise — focus on what's useful for future conversations