devkit is a professional, zero-dependency R development toolkit designed to streamline package management, environment setup, session auditing, and batch processing. It provides a suite of utilities to help developers maintain CRAN compliance and optimize their development workflow.
Install the released version of devkit from CRAN with:
install.packages("devkit")
Or install the development version from GitHub with:
# install.packages("devtools")
# devtools::install_github("zankrut20/devkit")
install.packages("devkit", repos = c('https://zankrut20.r-universe.dev', 'https://cloud.r-project.org'))
Once installed, simply load the library to access all utilities:
library(devkit)
# Example: Audit your package dependencies
audit_dependencies()
# Example: Clean up memory hogs
hunt_zombies()
audit_dependencies(): Verifies DESCRIPTION file vs actual code usage.remove_package(): Smart package removal with orphan dependency checking.remove_user_installed_packages(): Cleans all user-installed packages while preserving base/recommended ones.scan_dependencies(): Identifies unused packages in your session.sweep_memory(): Removes large objects and triggers garbage collection.hunt_zombies(): Cleans hidden temp files and orphaned graphics devices.sweep_temp_cache(): Flushes R session caches across the system.bootstrap_dev_env(): Installs and attaches core dev tools.manage_deprecation(): Scaffolds deprecation wrappers and refactors calls.setup_preflight(): Configures Git pre-commit hooks for safety.setup_sentinel(): Enables dual-logging for session reproducibility.audit_script(): Snapshots session state before and after script execution.detect_masking(): Resolves namespace conflicts and function masking.export_snapshot(): Exports attached packages to an installation script.dispatch_checkpoints(): Crash-resilient batch processing with recovery.loop_guardian(): Memory-safe iteration with RAM monitoring.network_diplomat(): Rate-limited network requests with exponential backoff.architect_release(): Automates version bumping and release notes.architect_vignette(): Scaffolds CRAN-compliant vignettes.scaffold_parallel(): Generates parallel processing boilerplate.scaffold_tests(): Generates testthat boilerplate.simulate_clean_room(): Verifies script reproducibility in a vanilla session.mask_identity(): Interactively anonymizes PII in datasets.benchmark_branches(): Compares performance across Git branches.dictate_dictionary(): Generates roxygen2 documentation for data frames.This package is licensed under the MIT License.
Contributions are welcome! Please see CONTRIBUTING.md for guidelines.