Skill

test-plan-score-test-function

Score generated test function code for completeness, quality, and convention adherence using a 5-criteria rubric. Use for validating generated test code quality before including in the final implementation.

npx claudepluginhub opendatahub-io/skills-registry --plugin test-plan

Configuration

User-invocable: No — Claude only
Model: sonnet
Context: Runs in forked subagent

Tool Access

This skill uses the workspace's default tool permissions.

Preview

Internal scorer sub-agent that evaluates generated test function code using a 5-criteria quality rubric. Forked by `test-plan.case-implement` in Step 5.6.

Supporting Assets

calibration/README.mdcalibration/good-cypress-test.cy.tscalibration/good-go-test.gocalibration/good-pytest-test.pycalibration/good-typescript-test.spec.tsxcalibration/poor-cypress-test.cy.tscalibration/poor-go-test.gocalibration/poor-pytest-test.pycalibration/poor-typescript-test.spec.tsxprompts/score-test-function.md

SKILL.md

Similar Skills

using-superpowers

185.1k

Mandates invoking relevant skills via tools before any response in coding sessions. Covers access, priorities, and adaptations for Claude Code, Copilot CLI, Gemini CLI.

3 files

superpowers

Stats

Stars0

Forks1

Last CommitMay 6, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

Test Function Quality Scorer

Internal scorer sub-agent that evaluates generated test function code using a 5-criteria quality rubric. Forked by test-plan.case-implement in Step 5.6.

Usage

This skill is not user-invocable. It is called by:

test-plan.case-implement (Step 5.6) for quality assurance of generated tests

Inputs

From arguments

Parse $ARGUMENTS to extract:

--test-code-file: Path to file containing generated test code
--tc-file: Path to TC-*.md specification
--conventions-file: Path to repository conventions markdown
--framework: Test framework (pytest, Go testing, Jest, etc.)
--output-file: Path where assessment should be written (for revision feedback)

Process

Step 1: Read Scoring Instructions

Read the detailed scoring rubric from ${CLAUDE_SKILL_DIR}/prompts/score-test-function.md.

Step 2: Execute Scoring

Apply the prompt with substitutions:

{TEST_CODE_FILE} = --test-code-file argument
{TC_FILE} = --tc-file argument
{CONVENTIONS_FILE} = --conventions-file argument
{FRAMEWORK} = --framework argument
{OUTPUT_FILE} = --output-file argument

The scoring rubric evaluates:

Coverage (0-2): All TC requirements implemented?
Assertions (0-2): Specific and meaningful?
Convention Adherence (0-2): Follows repo patterns?
Test Data (0-2): Uses realistic values from TC?
Code Quality (0-2): Clean, no excessive TODOs?

Step 3: Write Assessment to File

Write the structured markdown assessment to the --output-file path.

The output includes:

Per-criterion scores with issues
Total score (0-10)
Verdict (Ready/Good/Revise/Rework)
Coverage analysis (preconditions, steps, assertions implemented)
Specific issues found
Revision recommendations (if needed)

IMPORTANT: The orchestrating skill needs this file to:

Extract the verdict to decide if revision is needed
Extract specific issues to pass as feedback to the test generator
Keep an audit trail of quality assessments

What This Skill Does NOT Do

Does NOT modify the generated test code (only assesses it)
Does NOT trigger auto-revision (that's handled by the orchestrating skill)
Does NOT make the revision decision (returns verdict, orchestrating skill decides)
For scoring + auto-revision, the orchestrating skill (test-plan.case-implement) handles the revision loop

$ARGUMENTS