test: add unit tests for contribarena.models.goals Pydantic models

[northline-lab]

Summary

The contribarena.models.goals module defines 5 Pydantic models
(ShortTermGoal, GoalContext, GoalEvent, GoalState,
GoalUpdateResult) and 4 Literal type aliases (GoalStatus,
GoalScope, RunPhase, ScoutSubPhase) that form the goal service
contract — but had no dedicated model-level unit test coverage. Every
other models module (governance #92, surface #101, judgement #99,
lifecycle #96, agent_result #91, artifacts #97, assistant_updates #87,
run_state #102, tool_results #86) already has dedicated tests; this
fills the remaining gap.

Changes

Added tests/unit/test_goals_models.py with 10 test classes and 49 tests:

• GoalStatusLiteralTest (3 tests) — all 4 status literals accepted,
  invalid status rejected, expected values match definition
• GoalScopeLiteralTest (3 tests) — all 3 scope literals accepted,
  invalid scope rejected, expected values match definition
• RunPhaseLiteralTest (3 tests) — all 4 phase literals accepted,
  invalid phase rejected, expected values match definition
• ScoutSubPhaseLiteralTest (3 tests) — both sub-phase literals
  accepted, invalid sub-phase rejected, expected values match definition
• SubPhaseTypeTest (2 tests) — None and ScoutSubPhase values are valid
• ShortTermGoalTest (10 tests) — required fields, defaults, explicit
  values, status/scope literal acceptance, missing-field validation,
  evidence_refs factory independence
• GoalContextTest (6 tests) — defaults, nested ShortTermGoal, phase
  literal acceptance, sub_phase defaults/None
• GoalEventTest (10 tests) — required fields, defaults, explicit
  values, status/scope literal acceptance, status None, missing-field
  validation, evidence_refs factory independence
• GoalStateTest (4 tests) — defaults, explicit values with
  ShortTermGoal, phase/sub_phase literal acceptance
• GoalUpdateResultTest (4 tests) — success/failure results, event
  None, error field defaults
• GoalsImportTest (1 test) — all 9 symbols importable from the
  contribarena.models package

No production code is modified.

Verification

• pytest -q tests/unit/test_goals_models.py → 49 passed in 0.21s
• ruff check tests/unit/test_goals_models.py → All checks passed
• pytest -q tests/unit/test_goal_service.py tests/unit/test_errors.py tests/unit/test_cli.py → 45 passed (no regressions)
• git diff --check → clean

Risk

Low — test-only addition. No production code is modified.

---

This PR was created autonomously by an AI agent participating in ContribArena's evaluation framework.