tldr

Search tldr

By Arthur-Ficial in AI2 min read

apfel/demo at main · Arthur-Ficial/apfel · GitHub

A collection of practical shell scripts that use apfel and Apple Intelligence for offline, on-device developer assistance. The demos turn natural-language requests into commands, explanations, project summaries, naming suggestions, and system insights.

Share

apfel/demo at main · Arthur-Ficial/apfel · GitHub

Author: Arthur-Ficial | Published: Unknown | Generated: 2026-04-03 | Domain: github.com Tags: ‘#apple-intelligence’ ‘#on-device-ai’ ‘#cli’ ‘#shell’ ‘#developer-tools’ ‘#macos’

TLDR

The apfel demo directory provides shell utilities powered by Apple Intelligence through the apfel CLI, running locally on Apple Silicon Macs with no API keys or cloud dependency. The scripts cover common developer workflows—generating shell commands and pipelines, explaining errors, summarizing projects and Git history, suggesting names, inspecting ports, and humorously narrating system state. They are designed for small, fast interactions within the runtime-detected Foundation Models context window: 4,096 tokens on macOS 26 and 8,192 on macOS 27.

Key Takeaways

  • Offline natural-language shell assistance: cmd translates requests such as “find all .log files modified today” into executable shell commands; it can copy results or execute them after confirmation.
  • Pipeline generation: oneliner specializes in composing commands with tools such as awk, sed, find, xargs, sort, uniq, grep, jq, and more.
  • Project and repository context: wtd inspects directory metadata—including README files, common package manifests, Git branch, and the latest commit—to identify a project’s purpose, language, and likely build/run command; gitsum summarizes recent commits in plain English.
  • Developer explanation and naming tools: explain accepts commands, errors, code, clipboard content, or piped input; naming proposes conventional names for functions, variables, classes, and files.
  • Local-install caveat: Demo scripts are deliberately excluded from brew install apfel and make install because names such as cmd, port, and explain are too generic. Users can create apfel--prefixed symlinks in ~/.local/bin to avoid $PATH collisions.
  • Platform requirements: Requires apfel on PATH, Apple Intelligence enabled, macOS 26 or later, and Apple Silicon.

Images & Media

None present in article.

  • apfel repository — Source repository for the on-device Apple Intelligence CLI and OpenAI-compatible server.
  • Demo README — Documentation and usage examples for the demo scripts.
  • Latest demo-directory commit — Documents dynamically detected context windows, including 8,192 tokens on macOS 27.

Keep reading

GitHub - Andyyyy64/whichllm: Find the local LLM that actually runs and performs best on your hardware. Ranked by real, recency-aware benchmarks, not parameter count. One command, run it instantly. · GitHub

whichllm is a Python CLI that detects or simulates hardware, then recommends runnable local LLMs using benchmark quality, recency, memory fit, and estimated speed—not parameter count alone. It also supports hardware planning, upgrade comparisons, instant model chat, and scriptable JSON output.