MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 2 callersFunction_suggestions
(result: ValidationResult)
tests/validators/test_plugin_component_risk.py:158
↓ 2 callersFunction_summarize_case
(case: dict)
src/skillevaluator/evaluation/insights_judge.py:238
↓ 2 callersFunction_summarize_cases_for_feedback
One row per agent and dataset case listing every attempt per arm.
src/skillevaluator/evaluation/tier3_feedback_chat.py:831
↓ 2 callersFunction_summarize_checked
(signals: Sequence[Mapping[str, Any]], key: str)
src/skillevaluator/tier3/eval_core/plugin_signals.py:4925
↓ 2 callersFunction_summarize_dimensions
Project SkillEvaluator dimensions into a compact prompt-friendly shape.
src/skillevaluator/evaluation/insights_judge.py:195
↓ 2 callersFunction_summarize_trajectory
Return safe trajectory metadata without raw prompts, outputs, or arguments.
src/skillevaluator/tier3/harbor/collector.py:6673
↓ 2 callersFunction_summary_is_scoreable
(data: dict[str, Any], allow_missing_status: bool)
src/skillevaluator/tier3/commands.py:1734
↓ 2 callersFunction_summary_overall
(data: dict[str, Any])
src/skillevaluator/tier3/commands.py:1749
↓ 2 callersFunction_summary_score_policy
(data: dict[str, Any])
src/skillevaluator/tier3/commands.py:1753
↓ 2 callersFunction_summary_standard_metrics
Return the standard metric set one condition summary declares.
src/skillevaluator/tier3/commands.py:1758
↓ 2 callersFunction_surface
(report: dict)
tests/validators/test_skillspector_hook_surface.py:325
↓ 2 callersFunction_swap_after_first_entry_stat
(monkeypatch: pytest.MonkeyPatch, module: object, swap: Callable[[], None])
tests/test_harbor_secure_copy.py:669
↓ 2 callersFunction_symlink
(link: Path, target: Path, *, directory: bool = False)
tests/test_plugin_manifest.py:34
↓ 2 callersFunction_synthetic_trajectory_content_sha256
Hash redacted semantic content without source identifiers or references.
src/skillevaluator/tier3/harbor/collector.py:2202
↓ 2 callersFunction_table
(*headers: str | Text)
src/skillevaluator/tier3/case_summary.py:605
↓ 2 callersMethod_target_end
(self, i: int)
src/skillevaluator/evaluation/dataset_case_removal.py:295
↓ 2 callersFunction_target_runtime_source_roots
Return standard and nested-package runtime subtrees that output cannot replace.
src/skillevaluator/tier3/harbor/adapter.py:692
↓ 2 callersFunction_terminal
(payload, *, verbose)
tests/reporting/test_declarative_coverage.py:185
↓ 2 callersFunction_terminate_harbor_process_tree
Terminate the POSIX process group or Windows task tree and reap Harbor. On POSIX the group receives SIGTERM, then SIGKILL once the leader exits o
src/skillevaluator/tier3/harbor/runner.py:3843
↓ 2 callersFunction_text
(value: Any, field: str, *, optional: bool = False)
src/skillevaluator/reporting/retained.py:58
↓ 2 callersFunction_text
(value: Any, limit: int, scrub: _Scrubber)
src/skillevaluator/evaluation/tier3_feedback_chat.py:721
↓ 2 callersFunction_text
Keep each cell on one bounded line, without executable Markdown or mentions.
ci/comment_summary.py:20
↓ 2 callersFunction_text
(markup)
tests/reporting/test_report_readability.py:17
↓ 2 callersFunction_text
(fragment: str)
tests/reporting/test_finding_verification.py:36
↓ 2 callersFunction_text
(fragment)
tests/reporting/native_comparison_assertions.py:10
↓ 2 callersFunction_text_is_safe
(text: str)
src/skillevaluator/skillspector_provenance.py:101
↓ 2 callersFunction_three_agent_run
opencode scores best, claude-code finishes lower, and codex fails.
tests/reporting/test_tier3_completed_agent_docs.py:116
↓ 2 callersFunction_threshold
(value)
src/skillevaluator/deduplication/plugin/catalog_checks.py:60
↓ 2 callersMethod_tickFormatFunction
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callersFunction_tier1_fix
(result: ValidationResult, problem: str = "")
src/skillevaluator/reporting/console_ui.py:1331
↓ 2 callersFunction_tier1_quality_messages
( command: list[str], tmp_path: Path, monkeypatch: pytest.MonkeyPatch, *extra_args: str, )
tests/test_commands.py:1028
↓ 2 callersFunction_tier1_results
(path: str = "scripts/fetch.py", repeat: int = 1, suffix: str = "")
tests/reporting/browser_navigation_fixtures.py:37
↓ 2 callersFunction_tier1_scope_strings
Copy bounded scope evidence without narrowing its publication meaning. Reject malformed declarations rather than dropping them: absence has legac
src/skillevaluator/reporting/json_reporter.py:77
↓ 2 callersFunction_tier3_configured_metrics
Read retained metric identities; never substitute the template's default metric set.
src/skillevaluator/reporting/html.py:1106
↓ 2 callersFunction_tier3_neutral_archive
(status, *, complete=True, blocking=True)
tests/ci/test_feedback.py:1100
↓ 2 callersFunction_tier3_text_mapping
Normalize public label maps without silently overwriting aliases.
src/skillevaluator/reporting/html.py:545
↓ 2 callersFunction_timed_out_scan
A scan that kept one partial finding, then timed out.
tests/reporting/test_report_redesign.py:133
↓ 2 callersFunction_timeout_plugin
(tmp_path: Path)
tests/validators/test_skillspector_hook_surface.py:212
↓ 2 callersFunction_timestamp_ms
Parse an ISO timestamp with a UTC offset; naive local times are rejected.
src/skillevaluator/tier3/harbor/trajectory_timing.py:126
↓ 2 callersFunction_token_counter_names
Return the token counter names that ``value`` lists, in ``_USAGE_COUNTERS`` order. Anything that is not such a name is dropped, so a report trial
src/skillevaluator/tier3/harbor/report_data.py:717
↓ 2 callersFunction_tool_call_ref
(step_idx: int, tc: dict[str, Any], *, kind: str)
src/skillevaluator/tier3/eval_core/atif_helpers.py:681
↓ 2 callersFunction_tool_efficiency_args
(tool_call: dict[str, Any])
src/skillevaluator/tier3/eval_core/checks.py:8253
↓ 2 callersFunction_tool_efficiency_name
Lowercased tool name without a Codex namespace such as ``functions.``. MCP tool names keep every part, so ``mcp__tasks__task`` stays an MCP call.
src/skillevaluator/tier3/eval_core/checks.py:8289
↓ 2 callersFunction_tool_efficiency_target
The command, code or path a tool call acts on, as recorded.
src/skillevaluator/tier3/eval_core/checks.py:8258
↓ 2 callersFunction_tool_name_looks_like_write
(fn_lower: str)
src/skillevaluator/tier3/eval_core/atif_helpers.py:435
↓ 2 callersFunction_tool_result_text
Flatten a tool result; a non-text block, such as an MCP image, becomes a placeholder.
src/skillevaluator/tier3/harbor/nvidia_build_bridge.py:1551
↓ 2 callersFunction_tool_write_body
(args: dict[str, Any])
src/skillevaluator/tier3/eval_core/atif_helpers.py:409
↓ 2 callersFunction_top_field
The top-level manifest field a Claude Code report path names (``"<root>"`` for the whole file).
src/skillevaluator/validators/claude_plugin_validate.py:228
↓ 2 callersFunction_top_k
(value)
src/skillevaluator/deduplication/plugin/catalog_checks.py:55
↓ 2 callersFunction_top_level_directory_names
Return the one-segment directory names of a nested file, such as ``policies``. ``src/<package>/`` also yields the package name, which imports use
src/skillevaluator/validators/skillspector_scope.py:675
↓ 2 callersFunction_traj_with_verified_write_and_long_answer
A run whose one expected behavior is provable from the write alone. The final answer is long enough that the final-response limit cuts its mi
tests/tier3/test_judge_contracts.py:409
↓ 2 callersFunction_trajectory_step_merge_identity
Return semantic step content without collector-owned identity fields.
src/skillevaluator/tier3/harbor/collector.py:1785
↓ 2 callersFunction_trajectory_with_leaked_runtime_key
Trajectory whose tool call, tool output, and final response leak *key_values*.
tests/tier3/test_behavior_evidence.py:339
↓ 2 callersFunction_tree
(root: Path, files: dict)
tests/tier3/test_private_runtime_regressions.py:35
↓ 2 callersFunction_tree
(root: Path, files: dict[str, Any])
tests/deduplication/plugin/test_openclaw_profile.py:60
↓ 2 callersFunction_tree_with_link
(tmp_path: Path)
tests/utils/test_secure_fs_platform.py:546
↓ 2 callersFunction_trial
Return whether a directory is a Harbor trial and whether that trial failed.
ci/diagnostics.py:98
↓ 2 callersFunction_trial
(tmp_path: Path, plan, *, init: dict | None)
tests/tier3/test_plugin_execution.py:226
↓ 2 callersFunction_trial_activation_validation
Keep routing, task quality, workflow and resources at their observed scopes. Fallback reward rows do not prove a complete logical trajectory. Eve
src/skillevaluator/evaluation/tier3_report.py:3675
↓ 2 callersFunction_trial_coverage_truncated
(marker: dict[str, Any])
src/skillevaluator/reporting/dashboard_data.py:113
↓ 2 callersFunction_trial_evidence_ref
(ref: Any, reward: dict[str, Any], *, multi_trial: bool)
src/skillevaluator/tier3/harbor/report.py:698
↓ 2 callersFunction_trial_result
(name: str, case: str, **fields: Any)
tests/tier3/test_baseline_timeouts.py:89
↓ 2 callersFunction_trials_by_case
Group an agent's with-skill trial directories by the case each one ran.
src/skillevaluator/tier3/case_summary.py:1038
↓ 2 callersMethod_trusted_compose_client_environment
Build a host/Harbor baseline without main or sidecar target state.
src/skillevaluator/tier3/harbor/secure_docker_environment.py:3049
↓ 2 callersMethod_trusted_docker_client_environment
Build a host-only Docker CLI environment without Compose/task state.
src/skillevaluator/tier3/harbor/secure_docker_environment.py:1428
↓ 2 callersFunction_trusted_md_cell
(value: object)
src/skillevaluator/reporting/benchmark.py:2174
↓ 2 callersMethod_unique_finding_count
Count a skill's findings once per ``finding_id``, before display merging. ``issue_count`` keeps counting raw records for the JSON ``skills``
src/skillevaluator/reporting/html.py:2583
↓ 2 callersFunction_unique_findings
(findings: Iterable[Finding])
src/skillevaluator/plugin_inventory.py:216
↓ 2 callersFunction_unlinked_directory
Return ``root/parts`` when every component below *root* is a directory and none is a link.
src/skillevaluator/embedding/registry.py:742
↓ 2 callersFunction_unquoted_separator_index
Where an operand ends: at the first separator outside quotes. ``cat <<< "a; b"`` is one operand, so the ``;`` inside the quotes does not end
src/skillevaluator/tier3/eval_core/script_invocation.py:2316
↓ 2 callersMethod_unrecorded
(reasons: list[str])
tests/test_benchmark_source_provenance.py:912
↓ 2 callersFunction_unscanned_packaged_finding
HIGH for a component file a client loads from a folder the scans skip, below the folder it lists. Clients load nested component folders whatever
src/skillevaluator/plugin_paths.py:426
↓ 2 callersFunction_unscanned_path_finding
HIGH when a declared component lives in a folder that Tier 1 whole-tree scans skip. The security, secret, and Unicode scans prune evaluation outp
src/skillevaluator/plugin_paths.py:399
↓ 2 callersFunction_unscoreable_reward_reason
Return a bounded, redacted diagnostic for an unscoreable reward.
src/skillevaluator/tier3/harbor/collector.py:4530
↓ 2 callersFunction_untrusted_result
(backend)
tests/reporting/test_execution_backend_reporting.py:152
↓ 2 callersFunction_unverified_case
(entry_id: str, *, in_dataset: bool = True)
tests/reporting/test_tier3_token_usage_reports.py:1282
↓ 2 callersFunction_unwrap
The original command a census-wrapped shell-form hook handler runs.
tests/tier3/test_private_loose_source.py:105
↓ 2 callersMethod_update
(t=Date.now())
src/skillevaluator/reporting/assets/chart.umd.min.js:7
↓ 2 callersMethod_updateVisibility
(t,e,i)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callersMethod_upgrade_suggestion
(pkg_name: str, fix_versions: list[str])
src/skillevaluator/validators/dependencies.py:957
↓ 2 callersFunction_upsert
(api, *, collection, member, owned, body, receipt, current, method="patch", verified=None)
ci/feedback.py:266
↓ 2 callersFunction_url
(value: object)
ci/report_views.py:53
↓ 2 callersFunction_url_parts
``(scheme, host, effective port, normalized path)`` of an absolute URL, or ``None`` when unparseable. The URL is read the way Claude Code's HTTP
src/skillevaluator/plugin_component_risk.py:2679
↓ 2 callersFunction_urlopen_with_schema_fallback
Open URL with retry, falling back to prompt-only on confirmed schema capability errors.
src/skillevaluator/tier3/harbor/templates/eval.py:1942
↓ 2 callersFunction_usage
(raw: Any)
src/skillevaluator/tier3/plugin_evidence.py:107
↓ 2 callersFunction_use_anthropic
(monkeypatch: pytest.MonkeyPatch)
tests/validators/test_rubric_eval.py:1277
↓ 2 callersFunction_user_facing_reports
Render every Tier 1 user-facing report surface for redaction checks.
tests/validators/test_security.py:380
↓ 2 callersFunction_valid_container_reference
(reference: str, *, require_digest: bool)
scripts/ci/check_public_benchmarks.py:258
↓ 2 callersFunction_valid_evaluation_date
Return whether an ISO date exists on the calendar and is not future-dated.
scripts/ci/check_public_benchmarks.py:1852
↓ 2 callersFunction_valid_evidence
(evidence: object)
src/skillevaluator/evaluation/tier3_report.py:4193
↓ 2 callersFunction_validate
(target: Path, reports: Path, *, workers: int, extra_args: list[str])
tests/test_catalog_results.py:54
↓ 2 callersFunction_validate_agent_generated_outputs
(agent_dir: Path, output_root: Path)
src/skillevaluator/tier3/harbor/collector.py:294
↓ 2 callersFunction_validate_compose_host_passthrough
(value: object, *, allowed_env: set[str], field: str)
src/skillevaluator/tier3/harbor/adapter.py:3335
↓ 2 callersFunction_validate_custom_grader_file
(path: Path, rel_path: str, results: list[CheckResult])
src/skillevaluator/tier3/evals_spec.py:1197
↓ 2 callersFunction_validate_directory_components
Reject every existing link/reparse component in a lexical directory path.
src/skillevaluator/tier3/case_ids.py:106
↓ 2 callersFunction_validate_directory_metadata
(metadata: os.stat_result, path: Path)
src/skillevaluator/reporting/base.py:158
↓ 2 callersFunction_validate_evals_source_directory
Reject a link/reparse-point evaluator root before loading any assets.
src/skillevaluator/tier3/harbor/adapter.py:1876
↓ 2 callersFunction_validate_evaluator_parent_snapshot
(snapshot: tuple[tuple[Path, tuple], ...], *, label: str)
src/skillevaluator/tier3/harbor/adapter.py:1809
↓ 2 callersFunction_validate_existing_destination_file
(metadata: os.stat_result, destination: Path)
src/skillevaluator/tier3/harbor/secure_copy.py:1676
↓ 2 callersFunction_validate_file_metadata
(metadata: os.stat_result, path: Path)
src/skillevaluator/reporting/base.py:165
← previousnext →4,701–4,800 of 24,741, ranked by callers