Inputs & Outputs

tip

The Workflow Wizard lets you configure these inputs visually and generates the complete workflow YAML for you.

Inputs

Input	Required	Default	Description
github_token	Yes	${{ github.token }}	GitHub token for API access — issues, releases, and repository metadata. Not used for question generation
ai_provider	No	openrouter	AI provider. openrouter is the only supported value, so this can be omitted
ai_model	No	google/gemini-3.5-flash-lite	Model identifier in OpenRouter provider/model-name format. See openrouter.ai/models
ai_retry_max_attempts	No	5	Total number of attempts (initial + retries) when calling the AI provider. Retries are triggered by transient errors: 429 (rate limit), 500, 502, 503, 504, and network-level failures. Values below 1 are clamped to 1
ai_temperature	No	0.5	Controls the randomness of the AI's output (0.0 = fully deterministic, 1.0 = most random). Lower values produce more consistent questions; higher values produce more varied output
api_key	Yes		OpenRouter API key. Required — github_token cannot be used for question generation, and the action fails immediately if this is empty. Create one at openrouter.ai/keys
num_questions	No	20	Number of questions to generate (minimum 1, maximum 50). Values above 50 are automatically capped
include_answers	No	false	When true, each question is immediately followed by its answer labelled Answer: in the student-facing report — meaning the student sees the answers. This defeats the purpose of the assessment, which is for the student to work out the answers themselves. Leave this false in almost all cases. The instructor repository (when instructor_repo_token is configured) always includes answers regardless of this setting
exclude_pattern_overrides	No		Comma-separated entries to re-include files excluded by auto-detection or additional_exclude_patterns. Each entry can be an exact pattern (e.g. **/*.md) to re-include all files of that type, or a specific file path (e.g. README.md) to allow only that file through. Note: binary files are always skipped regardless of overrides
additional_exclude_patterns	No		Comma-separated globs for extra files to exclude on top of the auto-detected stack patterns. Use for assignment-specific files (starter code, fixtures, data files) that the auto-detected templates wouldn't cover
instructor_repo_token	No		Classroom 50 assignment repositories only. PAT with repo and workflow scopes and permission to create repositories in the same organisation. When provided, the action writes a private instructor-only assessment file (questions and answers) to a repository named {assignment-name}-grillmycode-instructor in the same organisation. The repository is created automatically on first run, and its quiz-generation workflow and README are refreshed on every run whenever they differ from the copies shipped with the action. The assignment name and the student's folder are read from the Classroom 50 repository name (<classroom>-<assignment>-<username>) and the repository's direct collaborators — see how the assignment and student are identified. Any other repository skips instructor delivery with a warning. Leave empty to disable instructor repository delivery
instructor_context	No		Instructor-specific instructions for this assignment. Injected into the system prompt and takes precedence over default behaviour. Supports multi-line instructions. When set, the AI also generates a one-sentence summary of the question focus, shown as an Instructor Note in the report header
assignment_context	No		Comma-separated file glob(s) read from the repository and injected into the AI prompt before instructor_context. Supported file types: plain text / source files (UTF-8), PDF (.pdf — text layer only), Microsoft Word (.doc/.docx — text only). If no files match, a workflow warning is emitted and the action continues without context. Example: "README.md, docs/brief.pdf, rubric.docx"
assignment_context_max_chars	No	20000	Maximum total characters read from all assignment_context files combined. Prevents large files from flooding the prompt. Values below 1 are clamped to 1
keep_comments	No	false	When false (default), inline and block comments are stripped before sending code to the AI. Set to "true" to preserve comments
fail_on_empty_assessment	No	false	When true, a run that finds nothing to assess fails instead of succeeding. Both causes (empty commit range; every changed file excluded) occur normally at assignment-accept time, so this is opt-in. The run summary explains the reason either way
include_initial_commit	No	false	When false (default), pins the diff base to the first commit so starter/template files are excluded. When true, uses the empty tree as the base so the initial commit's eligible files are included in the diff. Set true for Classroom 50 empty-repository assignments — see the Classroom 50 guide
skip_committers	No	github-actions[bot]	Comma-separated list of commit author names or email substrings. Leading bot commits after the base SHA are excluded from the diff. Classroom 50's accept-time setup commit is authored under the student's own identity, not a bot, so this input has no leading commit to match there (see Classroom 50); its .classroom50.yaml file is excluded by pattern instead. Set to '' to disable
base_sha	No		Override the base commit SHA
head_sha	No		Override the head commit SHA

Outputs

Output	Description
issue_url	URL of the created or updated assessment issue
issue_number	Number of the created or updated assessment issue
pdf_url	Download URL of the assessment PDF release asset. Empty if PDF generation failed
questions	The generated questions as a string (internal AI markers stripped; does not include the instructor note)
code_before_strip	Full code content of all assessed files before comment stripping
code_after_strip	Full code content of all assessed files after comment stripping

Using outputs in subsequent steps

Give the action step an id, then reference its outputs with steps.<id>.outputs.<name>:

- uses: NSCC-ITC-Assessment/GrillMyCode@v1
  id: assess
  with:
    github_token: ${{ secrets.GITHUB_TOKEN }}
    api_key: ${{ secrets.OPENROUTER_API_KEY }}

- name: Print issue link
  run: echo "Assessment issue ${{ steps.assess.outputs.issue_url }}"

- name: Print questions
  run: echo "${{ steps.assess.outputs.questions }}"

- name: Comment character count
  run: |
    echo "Code before stripping: $(echo "${{ steps.assess.outputs.code_before_strip }}" | wc -c) chars"
    echo "Code after stripping:  $(echo "${{ steps.assess.outputs.code_after_strip }}" | wc -c) chars"