autOScan

TUI tool for grading C lab submissions

autoscan-web.vercel.app

---

What It Does

autOScan helps you grade C submissions in bulk with policy-driven compile/run checks.

• Batch compile and classify submissions
• Detect banned function calls
• Run submissions with manual args/stdin or policy test cases
• Run multi-process labs with streaming output
• Compare output against expected files (diff view)
• Similarity detection across submissions
• AI-pattern detection against a configurable dictionary

---

Installation

Supported platforms: macOS (arm64), Linux (amd64)

Download binaries from Releases.

macOS:

chmod +x autoscan-darwin-arm64
./autoscan-darwin-arm64

Linux:

chmod +x autoscan-linux-amd64
./autoscan-linux-amd64

On first run, autOScan auto-installs to ~/.local/bin/autoscan and prompts for PATH updates if needed.

Build From Source

git clone https://github.com/autoscan-lab/autOScan.git
cd autOScan
make install

Requires: Go 1.22+, gcc

---

Quickstart

autoscan

1. Choose Run Grader.
2. Pick a policy.
3. Select the submissions folder.
4. Review results and run tests if needed.

---

Configuration

On first run, defaults are created at:

~/.config/autoscan/
├── policies/
├── libraries/
├── test_files/
├── expected_outputs/
├── banned.yaml
├── ai_dictionary.yaml
└── settings.yaml

Settings

• keep_binaries: keep compiled binaries for the Run tab
• max_workers: cap parallel compile workers (0 = all cores)
• plagiarism_*: similarity detection tuning
• ai_*: AI-pattern detection tuning

---

Policy Basics

Single-Process Example

name: "Lab 03"
compile:
  gcc: "gcc"
  flags: ["-Wall", "-Wextra", "-lpthread"]
  source_file: "lab03.c"
run:
  test_cases:
    - name: "Basic test"
      args: ["2", "3"]
      expected_exit: 0
library_files:
  - hospital.h
  - hospital.o
test_files:
  - input.txt

Multi-Process Example

name: "Lab 07 - Message Queues"
compile:
  gcc: "gcc"
  flags: ["-Wall"]
run:
  multi_process:
    enabled: true
    executables:
      - name: "Producer"
        source_file: "producer.c"
        args: ["queue1", "5"]
      - name: "Consumer"
        source_file: "consumer.c"
        args: ["queue1"]
        start_delay_ms: 100
    test_scenarios:
      - name: "No args"
        process_args:
          Producer: []
          Consumer: []
        expected_exits:
          Producer: 1
          Consumer: 1

Bundled Files

When added through the policy editor, files are copied into config directories so policies work from any working directory:

• library_files -> ~/.config/autoscan/libraries/
• test_files -> ~/.config/autoscan/test_files/
• expected outputs -> ~/.config/autoscan/expected_outputs/

---

Running Submissions

The Run tab supports:

• Manual execution with custom args/stdin
• Running policy test cases
• Multi-process execution with real-time output
• Expected output diff checks (expected_output_file / expected_outputs)

Requirement: enable keep_binaries before grading if you want to execute compiled programs in detail view.

---

License

MIT