obligation-test

Per-Obligation Test Derivation

Write one behavioral test per obligation with 6 hard quality rules enforced mechanically. Each obligation gets deep research (scenario + production code + test strategy) before any test code is written.

Args

/dev-practices-obligation-test UC-004-MAIN-01 UC-004-MAIN-02 UC-004-MAIN-03
/dev-practices-obligation-test UC-004 --count 10
/dev-practices-obligation-test UC-001-MAIN --count 15

Direct IDs: Space-separated obligation IDs. Each must be a behavioral obligation in your project's obligation documentation.

Prefix mode: A use-case prefix (e.g., UC-004, UC-001-MAIN). Auto-selects uncovered obligations from the allowlist matching the prefix. --count N limits how many (default: 10).

Protocol

Step 0: Resolve obligation IDs

If args look like a prefix (no trailing -NN sequence number):

python3 -c "
import json
al = json.loads(open('path/to/obligation_coverage_allowlist.json').read())
matches = sorted(oid for oid in al if oid.startswith('{prefix}'))
print(' '.join(matches[:N]))
print(f'Total matching: {len(matches)}, selected: {min(N, len(matches))}')
"

Store the resolved IDs for Step 1.

Step 1: Cook the molecule

python3 ${CLAUDE_PLUGIN_ROOT}/skills/dev-practices-execute/scripts/cook_formula.py \
  --formula ${CLAUDE_PLUGIN_ROOT}/skills/dev-practices-execute/formulas/obligation-test.yaml \
  --var "OBLIGATION_IDS={resolved_ids}" \
  --epic-title "Obligation tests: {prefix} batch N ({count} obligations)"

This creates: 1 setup atom + (4 atoms x N obligations) + 2 finalize atoms.

Step 2: Baseline (setup atom)

Run project quality command, record pass count, allowlist size, and coverage count in the epic notes. Close the baseline atom.

Step 3: Research all obligations (parallel)

All research atoms unblock after baseline. For each obligation:

1. Read the scenario from obligation documentation. Extract: Given/When/Then, business rule, priority, layer.

2. Translate Given/When/Then directly into the test:

   • Given -> test setup (fixtures, env configuration)
   • When -> action (call production function)
   • Then -> assertions (expected output/state)