Orchestrata Ultimate › Tests › Reference › agents › volks
Tests reference — tests/agents/volks
tests/agents/volks — 68 test modules, 1342 test functions (cumulative over subdirectories). Source tree: tests/agents/volks at commit 8a2cb42ad. Generated from the tests AST snapshot by scripts/gen-data-pages.py.
Prose pending.
Subdirectories
| Directory | Modules | Tests |
|---|---|---|
analyst/ |
5 | 55 |
architect/ |
7 | 110 |
codedirector/ |
15 | 408 |
compliance/ |
3 | 64 |
concierge/ |
3 | 20 |
documentor/ |
2 | 26 |
junior/ |
4 | 75 |
pm/ |
8 | 131 |
qa/ |
10 | 377 |
ux/ |
3 | 10 |
Modules
| Module | Tests | Covers |
|---|---|---|
test_base.py |
9 | 2 |
test_base_memory_fold.py |
4 | 6 |
test_base_observability.py |
8 | 4 |
test_base_tool_loop.py |
13 | 3 |
test_memory_context.py |
18 | 3 |
test_persona.py |
6 | 1 |
test_senior_dev.py |
8 | 5 |
Module details
tests/agents/volks/test_base.py
source · 9 tests · 129 lines
Tests for BaseAgent (volks base class).
Covers: agents.volks.base, core.packet/ (package)
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_base_agent_defaults |
23 | |
| function | test_base_agent_id_and_timestamps_are_unique_per_instance |
33 | |
| function | test_set_postman_and_set_kg_stores_private_refs |
39 | |
| function | test_send_raises_when_postman_not_set |
49 | |
| function | test_send_routes_via_postman_and_returns_response |
64 | |
| function | test_emit_event_builds_system_packet_and_routes |
92 | |
| function | test_process_raises_not_implemented |
106 | |
| function | test_get_output_display_empty |
120 | |
| function | test_get_output_display_joins_artifacts |
125 |
tests/agents/volks/test_base_memory_fold.py
source · 4 tests · 180 lines
Tests folding SPEC-010 memory recall into all BaseAgent LLM call paths.
Covers: agents.volks.base, agents.volks.memory_context, core.base.traces, llm.client, memory.memory_engine, core.protocols/ (package)
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_call_llm_real_path_folds_recalled_memory_into_graph_context |
50 | |
| function | test_call_reasoner_real_path_folds_recalled_memory_into_graph_context |
78 | |
| function | test_call_llm_with_tools_real_path_folds_recalled_memory_into_user_message |
106 | |
| function | test_all_three_paths_pass_through_cleanly_with_no_memory_registered |
137 |
tests/agents/volks/test_base_observability.py
source · 8 tests · 170 lines
Tests for BaseAgent as a node of the volk event mesh (Observer/Observable).
Covers: agents.volks.base, core.patterns.observer, postman.postman, core.packet/ (package)
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_emit_observation_delivers_event_with_all_keys |
51 | |
| function | test_emit_event_dual_emits_to_local_observer_and_cxp |
74 | |
| function | test_emit_event_without_postman_still_delivers_locally_then_raises |
101 | |
| function | test_base_agent_is_a_structural_async_observer |
114 | |
| function | test_one_volk_can_subscribe_to_another_with_override |
119 | |
| function | test_default_on_event_logs_at_debug |
135 | |
| function | test_unsubscribe_handle_stops_delivery |
146 | |
| function | test_raising_observer_does_not_block_others |
159 |
tests/agents/volks/test_base_tool_loop.py
source · 13 tests · 530 lines
Tests for the tool-EXECUTING branch of
BaseAgent._call_llm_with_tools.
Covers: agents.volks.base, llm.client, core.protocols/ (package)
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_tool_call_executes_against_real_executor_and_returns_final_response |
63 | A single real tool_call: executed, appended with the right shape, then |
| function | test_malformed_and_unknown_tool_calls_degrade_without_crashing |
130 | (d) A malformed arguments JSON string and an unknown tool name each |
| function | test_max_turns_guard_triggers_finalization_call_without_tools |
195 | (e) base.py's max_turns guard (default 10) tool-executing turns. If |
| function | test_custom_max_turns_is_honored |
275 | (F13, dogfood run-4) max_turns is a keyword parameter, not a |
| function | test_finalization_dsml_only_blob_returns_empty_and_logs_warning |
331 | (dogfood run-3e/F12) If DSML tool-call markup slips through Fix 1 and |
| function | test_finalization_dsml_plus_prose_returns_prose_only |
380 | (dogfood run-3e/F12) When the finalization response mixes real prose |
| function | test_tool_call_drives_codebase_module_deps_against_fixture_kb |
422 | Wiring proof (reports/proposal-prolog-next-impact.md §4): the real |
| class | TestStripDsmlToolMarkup |
497 | Direct unit tests of the pure _strip_dsml_tool_markup sanitizer |
| method | TestStripDsmlToolMarkup.test_pure_dsml_blob_returns_empty |
501 | |
| method | TestStripDsmlToolMarkup.test_dsml_plus_prose_returns_prose_only |
505 | |
| method | TestStripDsmlToolMarkup.test_clean_prose_is_unchanged |
512 | |
| method | TestStripDsmlToolMarkup.test_closed_marker_preserves_trailing_prose |
516 | A properly-closed <|...|> tag surrounded by prose on both |
| method | TestStripDsmlToolMarkup.test_unterminated_marker_strips_to_end_of_string |
523 | No closing "|>" anywhere -> the span runs to end-of-string, but |
| method | TestStripDsmlToolMarkup.test_empty_string_returns_empty_string |
529 |
tests/agents/volks/test_memory_context.py
source · 18 tests · 333 lines
Tests for the run-scoped memory-context provider (
memory.memory_context).
Covers: core.base.traces, memory.memory_context, memory.memory_engine
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_format_memories_from_objects_and_dicts |
27 | |
| function | test_format_memories_empty_returns_blank |
34 | |
| function | test_format_memories_respects_limit |
40 | |
| function | test_registry_roundtrip |
46 | |
| function | test_register_ignores_blank_thread_or_none_engine |
56 | |
| function | test_double_register_single_unregister_keeps_provider |
62 | P-H1: two binders on one thread (same-thread concurrent rounds) share |
| function | test_newest_engine_wins_while_counted |
77 | The newest engine passed to register_memory is always the active |
| function | test_balanced_unregister_drops_and_purges_caches |
93 | Recall/summary caches for a thread are purged only once the refcount |
| function | test_f4a_concurrent_rounds_recall_survives_first_unbind |
125 | F4a regression: round R1 and round R2 bind the same thread |
| function | test_recall_no_provider_or_empty_query_is_blank |
161 | |
| function | test_recall_with_real_engine_retrieves_stored_memory |
167 | |
| function | test_recall_caches_per_run_and_clears_on_unregister |
189 | |
| function | test_summarize_for_thread_with_real_engine |
218 | |
| function | test_summarize_for_thread_caches_and_clears_on_unregister |
239 | |
| function | test_summarize_for_thread_sync |
264 | |
| function | test_recall_sync |
285 | |
| function | test_recall_sync_respects_max_tokens |
306 | A small max_tokens cap truncates the rendered recall block. |
| function | test_summarize_for_thread_no_provider_is_empty |
331 |
tests/agents/volks/test_persona.py
source · 6 tests · 46 lines
Tests for the shared volk team-persona preamble (
agents.volks.persona).
Covers: agents.volks.persona
| Kind | Name | Line | Description |
|---|---|---|---|
| function | test_compose_prepends_preamble |
9 | |
| function | test_compose_empty_role_yields_preamble_only |
17 | |
| function | test_compose_is_idempotent |
22 | |
| function | test_preamble_conveys_team_cxp_and_status |
28 | |
| function | test_preamble_tells_volks_they_can_message_peers |
38 | |
| function | test_preamble_mandates_entities_used_section |
43 |
tests/agents/volks/test_senior_dev.py
source · 8 tests · 208 lines
Tests for agents.volks.senior_dev.
Covers: agents.tool_registry, llm.client, agents.volks.senior_dev/ (package), core.packet/ (package), core.protocols/ (package)
| Kind | Name | Line | Description |
|---|---|---|---|
| class | TestSeniorDevAgent |
60 | |
| method | TestSeniorDevAgent.test_init |
61 | |
| method | TestSeniorDevAgent.test_alias_is_u_sd |
67 | |
| method | TestSeniorDevAgent.test_constraints_are_per_instance |
78 | |
| method | TestSeniorDevAgent.test_parse_response_full |
88 | |
| method | TestSeniorDevAgent.test_parse_response_no_addendum |
105 | |
| method | TestSeniorDevAgent.test_parse_response_empty |
117 | |
| method | TestSeniorDevAgent.test_process |
124 | process() now drives its analysis through _call_llm_with_tools, |
| method | TestSeniorDevAgent.test_process_twice_does_not_leak_constraints_across_calls |
171 | Regression: _parse_response only APPENDS to self.constraints and |
Support files
Fixtures, package markers and helpers without test functions.