map-test-surface

name	map-test-surface
description	Enumerate every test file in a codebase, extract test subjects and assertions, detect the test framework in use, and emit test-surface.yaml. Used by test-engineer.
user-invocable	false
model	sonnet
allowed-tools	Read, Write, Grep, Glob, Bash
deprecated	true
deprecated_note	#434 ProductOS realignment — superseded by the command model; retained for Phase E reference, not installed

map-test-surface

Model-invocable skill for mapping the test surface of a codebase.

Purpose

Produce a complete inventory of existing tests: file locations, subjects (what they test), assertion counts, frameworks used. Emit test-surface.yaml — the basis for later blast-radius computation and coverage-gap analysis.

Previously authored inline by test-engineer.

Input

Field	Required	Description
project_root	yes	Codebase root
test_globs	optional	Override test-file globs (default auto-detect by language)
output_base	yes	Directory to write test-surface.yaml

Process

1. Detect test framework(s). From config (pytest.ini, jest.config, go.sum, etc.) and file naming (*_test.py, *.test.ts, *_test.go, spec/ dirs).

2. Glob test files. Language-specific defaults; user override via test_globs.

3. Per file, extract:

   • framework — pytest | unittest | jest | vitest | mocha | go-test | junit | ...
   • subjects — the module/class/function each test block exercises (derived from imports + test names)
   • test_count — number of test functions / it(...) / methods
   • assertion_count — approximate count via grep (assert, expect(, should)
   • tags — skip / xfail / slow / integration markers

4. Aggregate. Total test count, assertion count, per-framework counts, subjects covered.

5. Emit test-surface.yaml:

   project_root: "{project_root}"
   generated_at: "{ISO-8601}"
   frameworks: [ ... ]
   totals:
     test_files: {n}
     test_functions: {n}
     assertions: {n}
   files:
     - path: "{test file}"
       framework: "{name}"
       subjects: [ "{module/class/function under test}" ]
       test_count: {n}
       assertion_count: {n}
        [  ]
   
       
        [  ]
   

Output

test_surface_path: "{output_base}/test-surface.yaml"
test_file_count: {n}
status: written

Boundaries

• Read-only.
• Do not execute tests — that is quality-auditor's domain.
• Approximate assertion counts via grep are acceptable — deep AST analysis is out of scope.