MCPcopy Create free account

hub / github.com/HKUDS/Vibe-Trading / functions

Functions26,887 in github.com/HKUDS/Vibe-Trading

Methodtest_the_default_chooser_offers_only_usable_accounts_and_preselects_nothing
(self, monkeypatch)
agent/tests/test_cli_live.py:653
Methodtest_the_delivered_text_is_still_an_exact_prefix
`content` stays the document itself, so next_offset keeps its meaning.
agent/tests/test_agent_output_discipline.py:824
Functiontest_the_derivatives_desk_can_reach_an_options_chain
(presets)
agent/tests/test_preset_honesty.py:114
Methodtest_the_diff_itself_is_uncapped_so_counts_stay_truthful
Trimming belongs to the budget step; capping here would corrupt the totals.
agent/tests/test_institutional_holdings_tool.py:471
Functiontest_the_document_is_a_str_that_carries_the_fetch_evidence
()
agent/tests/test_paper_trade_5m_tool.py:207
Functiontest_the_documented_invariant_is_cvar_ge_var_and_not_var_ge_zero
()
agent/tests/quantlib/test_risk.py:147
Functiontest_the_error_message_lists_what_is_accepted
()
agent/tests/quantlib/test_options.py:483
Methodtest_the_error_names_every_currency_and_its_codes
(self)
agent/tests/test_composite_currency_guard.py:88
Functiontest_the_fallback_after_an_analysis_names_the_figures_not_prices
( two_runs: GroundingLedger, )
agent/tests/test_grounding_backtest_outputs.py:667
Methodtest_the_first_call_returns_an_outline
(self, tool: LoadSkillTool)
agent/tests/test_agent_output_discipline.py:806
Methodtest_the_first_page_reports_the_full_period_count
(self)
agent/tests/test_tool_result_paging.py:149
Methodtest_the_first_window_is_the_filing_season_itself
(self)
agent/tests/test_institutional_holdings_tool.py:1520
Methodtest_the_funding_span_of_a_period_bar
(self)
agent/tests/test_weekly_monthly_bars.py:185
Functiontest_the_gate_is_measurably_better_on_both_sides
The precision/recall claim in the section header, as an executable check.
agent/tests/test_research_papers_tool.py:821
Methodtest_the_gate_is_strict_about_encyclopaedic_figures_on_purpose
Where it is strict, it is strict knowingly. A definitional figure recited about an instrument the session never fetched is still a fi
agent/tests/test_agent_output_discipline.py:679
Functiontest_the_generated_engine_applies_a_one_sided_bound
The executed engine must filter on a one-sided bound, not just print it. Each side of a bound carries its own entry, for RSI as well as for the
agent/tests/test_shadow_codegen_condition_bounds.py:156
Functiontest_the_governance_block_is_present_even_when_empty
(tools)
agent/tests/test_sdm_governance_wiring.py:55
Methodtest_the_grounded_answer_passes_untouched
(self, tmp_path: Path)
agent/tests/test_agent_output_discipline.py:302
Functiontest_the_guard_can_actually_see_an_offender
The scan must fail on a padded call, not just on nothing existing.
agent/tests/test_skill_pct_change_guard.py:60
Functiontest_the_haircut_is_exactly_the_expected_maximum_under_the_null
(monkeypatch)
agent/tests/test_bench_multiple_testing.py:61
Functiontest_the_id_place_order_issued_can_be_cancelled
(monkeypatch, mod, Config, order_kwargs)
agent/tests/test_paper_sim_cancel_ids.py:40
Functiontest_the_indicator_tool_fetches_whole_periods
(monkeypatch)
agent/tests/test_weekly_monthly_bars.py:196
Methodtest_the_installed_adapter_renders_effort_as_output_config
Guards the assumption the whole change rests on. `reasoning_effort` is only worth forwarding because langchain-anthropic turns it int
agent/tests/test_llm_reasoning_effort.py:555
Functiontest_the_keep_list_caps_and_counts_the_rest
A long keep list stays one line: 24 values spelled, the rest counted.
agent/tests/test_grounding_release.py:2413
Functiontest_the_keep_list_is_absent_when_nothing_passed
No clean figure, no keep list: the section must not render empty.
agent/tests/test_grounding_release.py:2404
Methodtest_the_largest_skill_pages_in_a_sane_number_of_calls
(self, tool)
agent/tests/test_load_skill_paging.py:106
Functiontest_the_last_block_wins_and_earlier_fences_are_code
A draft may show code; only the trailing ``figures`` fence declares.
agent/tests/test_grounding_figures_contract.py:180
Methodtest_the_late_tail_is_only_paged_when_the_sample_is_short
(self)
agent/tests/test_institutional_holdings_tool.py:1532
Methodtest_the_latest_bar_keeps_its_traded_price
(self)
agent/tests/test_alpha_bench_qfq.py:66
Functiontest_the_layout_test_mock_states_the_packaged_version
The mocked sidebar version drifts silently otherwise. It is fixture data and nothing asserts on it, which is precisely why it sat at v0.1.13
agent/tests/test_release_version_consistency.py:84
Functiontest_the_lifespan_is_the_sole_production_start_call_site
``_start_scheduled_research_executor()`` is called from api_server only. A second production call site would break the single-writer invariant
agent/tests/test_scheduling_lifespan_owner.py:415
Methodtest_the_longest_price_frame_decides
A short hourly stub beside a long daily series does not switch the run.
agent/tests/test_cross_market_annualization.py:407
Methodtest_the_lookback_asks_for_each_year_at_most_once
(self)
agent/tests/test_etf_holdings_tool.py:650
Functiontest_the_manifest_records_the_tool_registry
(trace_dir)
agent/tests/test_run_manifest_wiring.py:53
Functiontest_the_marker_follows_each_lines_script_not_the_documents
The document-wide sweep picked ONE marker for every line it touched. A Chinese report containing an English table released "| First |(略※) |",
agent/tests/test_grounding_release.py:1870
Functiontest_the_marker_follows_the_script_of_the_text_it_cuts
A Chinese user gets English tables; the marker must match the clause, not the user.
agent/tests/test_grounding_release.py:1533
Functiontest_the_matcher_reads_the_endpoints_instruction
(message: str, expected: bool)
agent/tests/test_llm_responses_self_heal.py:326
Functiontest_the_matching_call_ref_validates_once_chosen
(tmp_path: Path)
agent/tests/test_grounding_container_field_refs.py:202
Methodtest_the_message_includes_the_response_body
(self)
agent/tests/test_mcp_client_adapter.py:848
Functiontest_the_model_is_remembered_so_the_next_call_skips_the_refusal
()
agent/tests/test_llm_responses_self_heal.py:218
Functiontest_the_module_string_never_reaches_importlib
(tool)
agent/tests/test_quantlib_tool.py:37
Functiontest_the_multiples_of_a_hand_computed_fund
()
agent/tests/quantlib/test_fundmath.py:372
Functiontest_the_note_states_which_direction_was_taken
A fall and the rise back between the same endpoints are different notes. The direction used to be inferred from the word beside the figure — 回撤
agent/tests/test_grounding_release.py:999
Methodtest_the_notice_survives_an_absurdly_small_budget
(self)
agent/tests/test_tool_result_paging.py:46
Functiontest_the_observed_range_includes_its_extremes
(tmp_path: Path, level: str, inside: bool)
agent/tests/test_grounding_role_hardening.py:371
Functiontest_the_old_fixed_threshold_evicts_the_same_run
Mutation side: the pre-fix 40K estimate clears filings mid-comparison.
agent/tests/test_context_budget.py:276
Functiontest_the_omitted_record_cannot_be_overwritten_by_a_metric
``_omitted`` names what the block dropped, so a metric may not claim it. Publishing a metric literally called ``_omitted`` would let one metric b
agent/tests/test_run_card_structured_metrics.py:178
Methodtest_the_originating_filing_supplies_the_fiscal_label
(self)
agent/tests/test_sec_period_frames.py:191
Functiontest_the_other_arguments_stay_required
Making session_id optional must not loosen the real inputs. goal_id/expected_goal_id are deliberately not in this set: they default to the cu
agent/tests/test_mcp_goal_session_contract.py:48
Methodtest_the_outline_envelope_respects_the_result_cap
(self, tool: LoadSkillTool)
agent/tests/test_agent_output_discipline.py:832
Methodtest_the_outline_labels_repeated_titles_with_their_path
The agent copies labels out of the outline, so the label must be addressable.
agent/tests/test_agent_output_discipline.py:1003
Methodtest_the_outline_names_every_section_and_its_size
(self, tool: LoadSkillTool)
agent/tests/test_agent_output_discipline.py:814
Functiontest_the_owner_survives_a_round_trip_through_the_store
(runtime_root)
agent/tests/test_principal_wiring.py:58
Functiontest_the_parser_is_lenient_about_form
§2: full-width pipes, run-on spacing and a missing ref are all accepted.
agent/tests/test_grounding_figures_contract.py:142
Methodtest_the_period_intervals_are_fetched_daily
(self)
agent/tests/test_weekly_monthly_bars.py:102
Functiontest_the_pinned_broker_count_matches_the_shipped_connectors
`EXPECTED_BROKERS` must equal the connectors the profile registry ships. Every other broker-count assertion compares a README against this pin, s
agent/tests/test_readme_counts.py:477
Functiontest_the_position_type_is_carried_through_never_interpreted
()
agent/tests/test_robinhood_mcp_mapping.py:121
Methodtest_the_prior_quarter_of_a_diff_is_resolved_the_same_way
Diffing against a superseded original would invent changes.
agent/tests/test_institutional_holdings_tool.py:1074
Methodtest_the_prompt_says_so_explicitly
(self)
agent/tests/test_agent_output_discipline.py:114
Methodtest_the_prompt_still_carries_all_five
(self)
agent/tests/test_agent_output_discipline.py:723
Methodtest_the_prompt_tells_the_model_which_ones_are_enforced
The model can only satisfy a gate it has been told about.
agent/tests/test_agent_output_discipline.py:735
Functiontest_the_prompt_text_itself_is_never_stored
(trace_dir)
agent/tests/test_run_manifest_wiring.py:61
Functiontest_the_raw_block_is_kept_in_the_artifact
Stripped from the answer, preserved in the audit record.
agent/tests/test_grounding_figures_contract.py:238
Functiontest_the_refusal_names_both_known_column_sets
#1170 asked for the refusal to name the sets it knows, not only the nearest one, so the error says what *is* supported rather than only what is wr
agent/tests/test_extraetf_reader.py:383
Functiontest_the_release_note_prints_a_large_price_in_full
``%g`` turns an index level into "1.23457e+06" six significant digits in.
agent/tests/test_grounding_release.py:1613
Functiontest_the_released_document_is_recorded_in_the_artifact
Every recheck on the release path is unrecorded, so the release itself is. Without this the artifact carried no evidence at all for the fail-clos
agent/tests/test_grounding_release.py:1928
Functiontest_the_released_document_is_validated_with_its_footnote
The footnote is answer text, so the gate must see it before release. The note carries the observed range and the symbols, and it used to be c
agent/tests/test_grounding_release.py:1641
Methodtest_the_report_is_handed_to_the_caller_for_the_run_card
(self)
agent/tests/test_cross_market_annualization.py:417
Methodtest_the_report_reaches_the_run_metrics_not_just_the_hook
A hook nobody merges into the result is the failure this guards.
agent/tests/test_china_futures_pricing_assumptions.py:126
Methodtest_the_report_reaches_the_run_metrics_not_just_the_hook
A hook nobody merges into the result is the failure this guards.
agent/tests/test_global_futures_pricing_assumptions.py:146
Methodtest_the_restated_value_wins_and_both_dates_are_visible
(self)
agent/tests/test_sec_period_frames.py:185
Functiontest_the_robinhood_live_profile_discloses_options_order_limit
()
agent/tests/test_robinhood_mcp_mapping.py:499
Methodtest_the_root_allowance_does_not_licence_an_untouched_ticker
Only the root a succeeding call actually passed in is licensed.
agent/tests/test_agent_output_discipline.py:611
Functiontest_the_same_figure_gets_the_same_verdict_under_any_prose
Prose cannot change a verdict, because no prose word is read. This replaces the price-word / level-word / observation-binder catalogues outri
agent/tests/test_grounding_release.py:1112
Functiontest_the_same_input_yields_byte_identical_documents
()
agent/tests/test_paper_trade_5m_golden.py:93
Functiontest_the_same_percent_declared_derived_passes
(tmp_path: Path)
agent/tests/test_grounding_role_hardening.py:362
Methodtest_the_same_quote_from_another_tool_is_also_accepted
Measured before this landed: rejected as numeric_claim_unavailable.
agent/tests/test_agent_output_discipline.py:192
Functiontest_the_same_served_tape_hashes_identically_twice
()
agent/tests/test_paper_trade_1d_fingerprint.py:79
Functiontest_the_same_span_is_finite_without_a_gap
Fixture check: the span above is only NaN because of the gap. Without this, an all-NaN output (a broken or warmup-only factor) would satisfy
agent/tests/factors/test_qlib158_cnt_nan_propagation.py:124
Functiontest_the_scan_sees_a_loguru_style_call
Guard the guard: an empty scan would pass vacuously.
agent/tests/test_channel_stdlib_logging.py:67
Methodtest_the_section_exists
(self)
agent/tests/test_agent_output_discipline.py:58
Functiontest_the_series_driven_waterfall_matches_the_hand_computed_split
100 called 2020-01-01, 150 distributed 2022-01-01, 8% simple pref, 20/100. 2020 is a leap year, so the accrual spans 731 days:: preferre
agent/tests/quantlib/test_fundmath.py:675
Functiontest_the_series_waterfall_distributes_realised_cash_by_default
()
agent/tests/quantlib/test_fundmath.py:715
Functiontest_the_series_waterfall_refuses_a_bare_list
()
agent/tests/quantlib/test_fundmath.py:733
Functiontest_the_settlement_currency_and_the_grounding_currency_follow_the_counter
()
agent/tests/test_hk_counter_currency.py:53
Functiontest_the_shortfall_is_neither_silently_filled_nor_silently_dropped
()
agent/tests/test_factor_costs.py:54
Functiontest_the_skill_category_table_matches_the_frontmatter
Each category row must carry the number of skills declaring it.
agent/tests/test_readme_counts.py:532
Functiontest_the_skill_coverage_limit_is_stated_not_implied
(trace_dir)
agent/tests/test_run_manifest_wiring.py:110
Functiontest_the_slash_table_is_the_router
The documented TUI commands must be exactly the ones the router resolves. This table had drifted furthest of anything in the file: it listed seve
agent/tests/test_readme_counts.py:382
Methodtest_the_smile_vol_differs_from_the_flat_vol_for_this_leg
(self)
agent/tests/test_options_smile_consistency.py:179
Functiontest_the_spanish_var_report_from_1418_grounds_once_its_confidence_is_declared
(tmp_path: Path)
agent/tests/test_grounding_role_hardening.py:1127
Methodtest_the_standing_filing_is_the_restatement_not_the_original
Downstream has to be able to tell which document the answer stands on.
agent/tests/test_institutional_holdings_tool.py:907
Functiontest_the_stream_path_retries_before_any_chunk_was_emitted
()
agent/tests/test_llm_responses_self_heal.py:246
Functiontest_the_subset_validator_defers_the_v12_version_gate_to_the_wire_builder
()
agent/tests/test_paper_trade_v12_schema.py:384
Methodtest_the_suggested_qualified_path_actually_resolves
(self, case_repeat: LoadSkillTool)
agent/tests/test_agent_output_discipline.py:1050
Functiontest_the_sweep_cuts_a_bare_restatement_of_a_cut_figure
The footnote's count is where the figure no longer appears. "37%" is a measurement and is cut at its span; "37 个百分点" is a bare integer the sh
agent/tests/test_grounding_release.py:2033
Functiontest_the_sweep_does_not_key_on_a_single_digit
Cutting "5%" must not take every "5" on the page; two digits still sweep (37 above).
agent/tests/test_grounding_release.py:2349
Functiontest_the_sweep_is_inert_without_collaborators
An install with no channel runtime must not be changed by any of this.
agent/tests/test_scheduled_delivery_outbox.py:211
Functiontest_the_sweep_leaves_a_bare_integer_with_different_digits
Cutting "0.95" must not take the "95" of an unrelated count with it.
agent/tests/test_grounding_release.py:2053
Functiontest_the_sweep_never_cuts_a_figure_the_gate_checked
Only bare integers are swept; a decimal sharing the cut figure's digits stays. "1.171%" is cut, and the observed close "1.171 元" carries the same
agent/tests/test_grounding_release.py:2068
← previousnext →25,201–25,300 of 26,887, ranked by callers