Task configuration

By default, task configuration is not required in assignment. But by its implementation, you are able to execute post-processing tasks after the assignment is submitted by student.

Example configuration file

This example configuration is only for demonstration purposes. You can find more details about each pipeline type in separate documentation tabs.

config.yml

# https://github.com/mrlvsb/kelvin/blob/master/README.pipeline.md
# You can also use CTRL+Space for autocompleting

# Specify the queue where the evaluation will be processed
queue: "evaluator"

# Max time for whole evaluation process
timeout: "15m"

# TODO: Add example tests
tests:

# Define the evaluation pipeline
pipeline:
  # Compile submitted source codes with gcc
  - type: gcc
    flags: -Wall -Wextra -g -fsanitize=address -lm -Wno-unused-variable

  # Runs clang-tidy static analysis and add warnings as comments
  - type: clang-tidy
    checks: '*'

  # Run configured tests
  - type: tests

  # Automatically propose points from the test results
  - type: auto_grader
    propose: true

# Manage resource limits
limits:
  fsize: 5M     # max filesize
  wall-time: 5  # max 5 seconds per test
  cg-mem: 5M    # max 5MB of memory usage

# Enable asynchronous tasks after evaluation
async:
  # LLM is module for automatic feedback generation using Large Language Models
  llm:
    enabled: false
    language: en
    model: gpt-4                      # Overwrite default model
    prompt_name: strict_code_review   # Use custom prompt