forked from HAL9000/cleveragents-core
freemo
051ee7c290
Added 52 new .feature files and corresponding _steps.py files targeting previously uncovered code paths in the following areas: - TUI layer: app, commands, persona (state/schema/registry), widgets, input (shell_exec, reference_parser) - Application services: plan lifecycle/service/executor, session, project, repo indexing, correction, checkpoint, actor, llm_actors, strategy coordinator, resource file watcher, service retry wiring - CLI commands: session, resource, repl, plan, db, automation_profile - Domain models: retry_policy, resource_type, cost_budget, docker_compose_analyzer, detail_level, _sql_string_aware, _postgresql_helpers - Core: circuit_breaker, retry_service_patterns - Infrastructure: repositories, transaction_sandbox, strategy_registry, plugins/loader, container - Config: settings - Agents: plan_generation, context_analysis, auto_debug - A2A: facade All new tests follow the Behave/Gherkin BDD standard. Resolved step definition collisions with unique prefixes. Fixed Alembic fileConfig logger disabling issue (disable_existing_loggers=False). ISSUES CLOSED: #1068
179 lines
9.5 KiB
Gherkin
179 lines
9.5 KiB
Gherkin
Feature: Plan Lifecycle Service coverage boost round 2
|
|
As a developer
|
|
I want to exercise remaining uncovered code paths in PlanLifecycleService
|
|
So that code coverage improves beyond 88.3%
|
|
|
|
Background:
|
|
Given I have a fresh plan lifecycle service for coverage boost r2
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 263-276: _consult_error_patterns — guidance injection path
|
|
# When error_pattern_service is configured and returns guidance,
|
|
# the formatted guidance is stored in plan.error_details.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: execute_plan injects preventive guidance from error pattern service
|
|
Given a plan lifecycle service with a configured error pattern service for r2
|
|
And an action "local/pattern-check" exists for coverage boost r2
|
|
And a plan advanced to strategize complete for coverage boost r2
|
|
When I execute the plan so error patterns are consulted
|
|
Then the plan error_details should contain preventive_guidance
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 277-282: _consult_error_patterns — exception handler path
|
|
# When error_pattern_service.match_patterns raises, the exception
|
|
# is caught and logged without propagating.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: execute_plan swallows error pattern service exceptions
|
|
Given a plan lifecycle service with a failing error pattern service for r2
|
|
And an action "local/pattern-fail" exists for coverage boost r2
|
|
And a plan advanced to strategize complete for coverage boost r2
|
|
When I execute the plan and error pattern consultation fails
|
|
Then the plan should still transition to execute phase without error
|
|
|
|
# ---------------------------------------------------------------
|
|
# Line 339: save_plan — public persistence convenience method
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: save_plan persists plan updates in memory mode
|
|
Given an action "local/save-test" exists for coverage boost r2
|
|
And a plan created from "local/save-test" for coverage boost r2
|
|
When I call save_plan on the plan
|
|
Then save_plan completes without error
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 416-422: _resolve_actor_registry_entry exception handler
|
|
# When the actor repository lookup raises, the exception is caught
|
|
# and None is returned.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: Actor registry resolution returns None when repository raises
|
|
Given a plan lifecycle service with a mock unit of work that fails actor lookup
|
|
When I resolve an actor registry entry for "local/broken-actor"
|
|
Then the resolved actor entry should be None
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 778-784: use_action — event_bus.emit exception handler
|
|
# When event_bus.emit raises during use_action, the exception
|
|
# is caught and the plan is still returned.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: use_action catches event bus emission failure
|
|
Given a plan lifecycle service with a failing event bus for r2
|
|
And an action "local/event-fail-use" exists for coverage boost r2
|
|
When I use action "local/event-fail-use" and event bus emit fails
|
|
Then the plan should still be created in strategize phase
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 847-850: list_plans — DatabaseError fallback to in-memory
|
|
# When the persistence layer raises DatabaseError on list_all(),
|
|
# the service falls back to in-memory cache.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: list_plans falls back to in-memory cache on DatabaseError
|
|
Given a plan lifecycle service with a mock unit of work that fails list_all
|
|
And an in-memory plan exists for the fallback test
|
|
When I call list_plans on the service with the broken persistence
|
|
Then the result should contain the in-memory plan
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 909-913: start_strategize — action loading from persistence
|
|
# When the plan's action_name is not in the _actions cache but is
|
|
# available via persistence, the service loads it.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: start_strategize loads action from persistence when not in cache
|
|
Given a plan lifecycle service with a persisted action for cache miss test
|
|
And a plan whose action is removed from the in-memory cache
|
|
When I start strategize on the plan with the cleared cache
|
|
Then start_strategize should complete successfully
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1070-1080: execute_plan — event_bus.emit exception handler
|
|
# When event_bus.emit raises during execute_plan phase transition,
|
|
# the exception is caught and the plan is still returned.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: execute_plan catches event bus emission failure on phase change
|
|
Given a plan lifecycle service with a failing event bus for r2
|
|
And an action "local/event-fail-exec" exists for coverage boost r2
|
|
And a plan advanced to strategize complete for coverage boost r2
|
|
When I execute the plan and event bus emit fails during phase change
|
|
Then the plan should still transition to execute phase
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1256-1276: complete_apply — event_bus emit success path
|
|
# Exercises the event_bus.emit call during complete_apply when
|
|
# the bus does NOT raise (emitting PLAN_APPLIED event).
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: complete_apply emits PLAN_APPLIED event via event bus
|
|
Given a plan lifecycle service with a recording event bus for r2
|
|
And an action "local/apply-event" exists for coverage boost r2
|
|
And a plan advanced to apply processing for coverage boost r2
|
|
When I complete apply on the plan with event bus
|
|
Then the event bus should have recorded a PLAN_APPLIED event
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1277-1282: complete_apply — event_bus emit exception path
|
|
# When event_bus.emit raises during complete_apply, the exception
|
|
# is caught and the plan is still returned.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: complete_apply catches event bus emission failure
|
|
Given a plan lifecycle service with a failing event bus for r2
|
|
And an action "local/apply-event-fail" exists for coverage boost r2
|
|
And a plan advanced to apply processing for coverage boost r2
|
|
When I complete apply on the plan and event bus emit fails
|
|
Then the plan should still be in applied state
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1366-1378: cancel_plan — event_bus emit success path
|
|
# Exercises the event_bus.emit call during cancel_plan.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: cancel_plan emits PLAN_CANCELLED event via event bus
|
|
Given a plan lifecycle service with a recording event bus for r2
|
|
And an action "local/cancel-event" exists for coverage boost r2
|
|
And a plan created from "local/cancel-event" for coverage boost r2
|
|
When I cancel the plan with event bus recording
|
|
Then the event bus should have recorded a PLAN_CANCELLED event
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1382-1388: cancel_plan — event_bus emit exception path
|
|
# When event_bus.emit raises during cancel_plan, the exception
|
|
# is caught and the plan is still returned.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: cancel_plan catches event bus emission failure
|
|
Given a plan lifecycle service with a failing event bus for r2
|
|
And an action "local/cancel-event-fail" exists for coverage boost r2
|
|
And a plan created from "local/cancel-event-fail" for coverage boost r2
|
|
When I cancel the plan and event bus emit fails
|
|
Then the plan should still be in cancelled state
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1417-1421: _cleanup_devcontainers — success path
|
|
# When CleanupService.stop_active_devcontainers returns stopped IDs.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: cleanup devcontainers logs stopped containers on plan completion
|
|
Given a plan lifecycle service with mocked devcontainer cleanup for r2
|
|
And an action "local/cleanup-ok" exists for coverage boost r2
|
|
And a plan advanced to apply processing for coverage boost r2
|
|
When I complete apply and devcontainer cleanup returns stopped containers
|
|
Then the plan should be applied and cleanup should have been called
|
|
|
|
# ---------------------------------------------------------------
|
|
# Lines 1422-1426: _cleanup_devcontainers — exception path
|
|
# When stop_active_devcontainers raises, the exception is caught.
|
|
# ---------------------------------------------------------------
|
|
|
|
Scenario: cleanup devcontainers catches exceptions without propagating
|
|
Given a plan lifecycle service with failing devcontainer cleanup for r2
|
|
And an action "local/cleanup-fail" exists for coverage boost r2
|
|
And a plan advanced to apply processing for coverage boost r2
|
|
When I complete apply and devcontainer cleanup raises an exception
|
|
Then the plan should still be applied despite cleanup failure
|