MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 3 callersMethod_run_strict
( self, command: list[str], run_root: Path, *, allow_net: bool = False
tests/test_harbor_local_sandbox.py:1102
↓ 3 callersFunction_runs
(job: dict[str, Any])
tests/test_ci_workflows.py:80
↓ 3 callersFunction_safe_host_homes
Return canonical environment and passwd homes, or fail closed.
src/skillevaluator/tier3/harbor/local_sandbox.py:651
↓ 3 callersFunction_safe_remove_fallback
(path: Path)
src/skillevaluator/tier3/harbor/secure_copy.py:1685
↓ 3 callersFunction_safe_text
(value: object)
src/skillevaluator/publication_text.py:329
↓ 3 callersFunction_safe_text
(value: Any, limit: int = 300)
src/skillevaluator/tier3/plugin_evidence.py:58
↓ 3 callersFunction_same_directory
Recognize actual filesystem aliases without folding distinct authored names.
src/skillevaluator/tier3/plugin_prepare.py:362
↓ 3 callersFunction_scalar
(value, label, max_chars, *, allow_empty=False)
src/skillevaluator/deduplication/plugin/llm_analyzer.py:31
↓ 3 callersFunction_scan_case
(directory, case)
tests/ci/test_smoke_image.py:84
↓ 3 callersMethod_schedule_creation_cleanup
( self, creation: asyncio.Task[asyncio.subprocess.Process], *, secret_values:
src/skillevaluator/tier3/harbor/local_environment.py:724
↓ 3 callersFunction_scopes
(usage: dict[str, Any], agent: str = "codex")
tests/tier3/test_token_usage.py:76
↓ 3 callersFunction_score_definition_for_policy
(metrics: tuple[str, ...], policy: str)
src/skillevaluator/tier3/harbor/collector.py:5774
↓ 3 callersFunction_secure_exec_arguments
Put env names on argv and every value in the child process env.
src/skillevaluator/tier3/harbor/secure_docker_environment.py:615
↓ 3 callersFunction_security_normpath
(path)
src/skillevaluator/tier3/eval_core/plugin_security.py:2366
↓ 3 callersFunction_security_within
``path`` is ``entry`` or inside it, one component at a time; a glob component of ``path`` matches too.
src/skillevaluator/tier3/eval_core/plugin_security.py:2449
↓ 3 callersFunction_selected_publication_target
Select one canonical target without accepting anonymous policy claims.
src/skillevaluator/reporting/base.py:945
↓ 3 callersFunction_selected_record
(path: Path, root: Path)
tests/tier3/test_decoys.py:54
↓ 3 callersMethod_send_json
(self, status: int, body: dict[str, Any])
src/skillevaluator/tier3/harbor/nvidia_build_bridge.py:287
↓ 3 callersFunction_sensitive_path_match
Return the longest entry of *paths* that *text* references, as written or normalized. The longest entry wins, so a write to /etc/profile.d/npm.sh
src/skillevaluator/tier3/eval_core/checks.py:938
↓ 3 callersFunction_server
(statuses, tls_context=None)
tests/test_retry_provider_transport.py:138
↓ 3 callersFunction_server_spellings
Casefolded spellings of a declared server name inside harness tool names, exact spelling first. Claude Code and OpenCode replace characters outsi
src/skillevaluator/tier3/eval_core/plugin_signals.py:2108
↓ 3 callersFunction_session_line
(call_id: str, server: str, tool: str, args: dict[str, Any])
tests/tier3/test_f90_mcp_execution.py:152
↓ 3 callersFunction_set_cached_tokens
(run: Path, trial: str, cached: object)
tests/reporting/test_tier3_token_usage_reports.py:250
↓ 3 callersFunction_settings_variant
(case: str)
tests/validators/test_skillspector_hook_surface.py:236
↓ 3 callersFunction_setup_fix
One item for checks that could not run because a scanner is missing or failed to download.
src/skillevaluator/reporting/console_ui.py:1347
↓ 3 callersFunction_setup_only_change
(env, workspace, git)
tests/ci/test_run_tier1_image.py:305
↓ 3 callersFunction_sha
(value)
ci/benchmark_publish.py:34
↓ 3 callersFunction_shell_inline
``(runs a -c program, the program)`` for a shell argv (``tokens[0]`` is the shell). Reads ``sh -c '<program>'``, combined flags (``bash -lc``, ``
src/skillevaluator/validators/mcp_static.py:1815
↓ 3 callersFunction_shell_substitution_payloads
Extract active command/process substitutions without evaluating shell text.
src/skillevaluator/tier3/eval_core/checks.py:1817
↓ 3 callersFunction_side_unavailable
Every alternative names a component type this arm cannot carry (see :func:`_ref_unavailable`).
src/skillevaluator/tier3/eval_core/plugin_signals.py:4243
↓ 3 callersFunction_signal_process_tree
(process: asyncio.subprocess.Process, value: signal.Signals)
src/skillevaluator/tier3/harbor/secure_docker_environment.py:670
↓ 3 callersFunction_single_skill
(root: Path, name: str)
tests/embedding/test_collection_coverage.py:179
↓ 3 callersFunction_skill
(name: str)
tests/test_openclaw_plugin_smoke.py:29
↓ 3 callersFunction_skill
(directory: Path)
tests/validators/test_plugin_dependencies.py:46
↓ 3 callersFunction_skill
(frontmatter: str)
tests/validators/test_plugin_static_risk_review_fixes.py:84
↓ 3 callersFunction_skill_name
(skill_path: Path, data: dict[str, Any])
src/skillevaluator/harness/overlay.py:243
↓ 3 callersFunction_skill_names
(source: NativePluginSource)
src/skillevaluator/tier3/plugin_native.py:1248
↓ 3 callersFunction_skill_names
(report: dict)
tests/cli/test_validate_manifest_file.py:37
↓ 3 callersFunction_skill_revision
(skill: dict[str, Any])
src/skillevaluator/tier3/version_comparison.py:41
↓ 3 callersFunction_skillspector_ledger_exception_is_hook_surface_limitation
Whether a ledger row is SkillSpector's limitation for hooks it parsed but could not model. The row must be a nonfatal, whole-file ``partial`` ``o
src/skillevaluator/validators/security.py:2776
↓ 3 callersFunction_skillspector_llm_env
Return the LLM-stage SkillSpector environment and the bridged SkillEvaluator provider.
src/skillevaluator/validators/security.py:769
↓ 3 callersFunction_skillspector_missing_reference_notes
Warnings for a ``partial`` analysis whose gaps are references to files not bundled, else ``None``. SkillSpector records a path-like reference in
src/skillevaluator/validators/security.py:2487
↓ 3 callersFunction_skillspector_reference_row_is_advisory
Whether a ledger row only records a reference to a file the skill does not bundle. The row must be a nonfatal ``partial`` row in phase ``referenc
src/skillevaluator/validators/security.py:1790
↓ 3 callersFunction_skillspector_report_path_key
A SkillSpector report *path* that is relative POSIX text, else ``None``. The path must be text that is safe to pass to OS path APIs, with no
src/skillevaluator/validators/security.py:1731
↓ 3 callersFunction_skillspector_scoring_match_identity
Return the producer's finding identity used for report compaction.
src/skillevaluator/validators/security.py:1128
↓ 3 callersFunction_skip_options
Index of the first positional token, skipping options and the values of ``value_flags``.
src/skillevaluator/validators/mcp_static.py:2035
↓ 3 callersFunction_skip_wrapper_options
Advance past one wrapper's own options using its grammar. Returns ``(_WRAPPER_OK, index)`` at the wrapped command, ``_WRAPPER_NONE`` when a h
src/skillevaluator/tier3/eval_core/script_invocation.py:2625
↓ 3 callersFunction_skipped
How many items the trials skipped because this arm cannot carry their component type (scored or not).
src/skillevaluator/tier3/eval_core/plugin_signals.py:4869
↓ 3 callersFunction_skipped_tier2
(message: str)
tests/test_commands.py:3595
↓ 3 callersFunction_source_snapshot
Read authored evaluator inputs only through the engine's secure snapshot.
src/skillevaluator/tier3/workflow.py:67
↓ 3 callersFunction_split_agent_model_states
Split agent records on commas outside model code spans.
scripts/ci/check_public_benchmarks.py:1109
↓ 3 callersFunction_split_by_headings
Split text at markdown heading boundaries, keeping headings with their content.
src/skillevaluator/embedding/client.py:369
↓ 3 callersFunction_stage_relocated_member
(tmp_path: Path, *, authored_skills_directory: bool = False)
tests/tier3/test_private_loose_dependency_scaffold_review.py:15
↓ 3 callersFunction_staged_environment_env
(task: Path)
tests/test_harbor_claude_background_tasks.py:43
↓ 3 callersFunction_stages
Instructions per stage. The unnamed final stage is the shipped runtime.
tests/ci/test_image_build.py:23
↓ 3 callersFunction_static
(version=None)
tests/reporting/test_report_version_placement.py:28
↓ 3 callersFunction_static_string
(node: ast.AST, *, depth: int = 0)
scripts/check_oss_boundary.py:290
↓ 3 callersFunction_stderr_shaped
(raw: io.RawIOBase)
tests/test_tier3_progress.py:2115
↓ 3 callersFunction_step_number_from_url
(url: str)
src/skillevaluator/evaluation/tier3_report.py:3959
↓ 3 callersMethod_stop_raw_service_containers
( self, service: str, *, remove: bool, require_existing: bool, )
src/skillevaluator/tier3/harbor/secure_docker_environment.py:1723
↓ 3 callersFunction_string_field
(frontmatter: dict[str, Any], *keys: str)
src/skillevaluator/plugin_component_risk.py:768
↓ 3 callersFunction_string_values
String values inside a JSON object or list (bounded, depth first), never key names.
src/skillevaluator/tier3/eval_core/plugin_signals.py:3946
↓ 3 callersFunction_strip_arm_suffix
Remove SkillEvaluator dual-arm suffixes (-with-skill, -without-skill) from an identifier.
src/skillevaluator/tier3/harbor/collector.py:4584
↓ 3 callersFunction_strip_link_target
(raw_target: str)
src/skillevaluator/tier3/harbor/adapter.py:1080
↓ 3 callersFunction_structural_attempt_ordinal
Read an ordinal only from runner-owned or exact legacy selector structure.
src/skillevaluator/tier3/harbor/collector.py:4267
↓ 3 callersFunction_stub_harbor_runner
Stub Harbor runner dependencies and capture the staged launch payload.
tests/test_harbor_task_toml_retry_env.py:67
↓ 3 callersFunction_subcommand
``(subcommand, tokens after it)`` of a runner front end, skipping its global options.
src/skillevaluator/validators/mcp_static.py:2050
↓ 3 callersMethod_suffixes
(self, component_type: str)
src/skillevaluator/plugin_discovery.py:947
↓ 3 callersFunction_summarize_evaluators
Drop any unknown evaluator fields and keep core scores.
src/skillevaluator/evaluation/insights_judge.py:213
↓ 3 callersFunction_supported_env_modes
()
src/skillevaluator/tier3/harbor/runner.py:1403
↓ 3 callersFunction_supports_custom_temperature
(model)
src/skillevaluator/tier3/harbor/templates/eval.py:1285
↓ 3 callersFunction_symlink_or_skip
(link: Path, target: Path)
tests/test_results_location.py:450
↓ 3 callersFunction_target_host
(url: str)
src/skillevaluator/tier3/mcp_proof.py:172
↓ 3 callersFunction_task_selector_from_harbor_result
Return one consistent staged selector, never ``[task].name``.
src/skillevaluator/tier3/harbor/collector.py:4128
↓ 3 callersFunction_template
()
tests/test_transport_rules.py:24
↓ 3 callersFunction_template_outputs
(tmp_path)
tests/tier3/test_metric_applicability_producer.py:333
↓ 3 callersFunction_template_task
The task template prompts ask for, and the outcome their graded text quotes. Both come from the description. When it reads, opens, loads, consult
src/skillevaluator/tier3/generate_dataset.py:470
↓ 3 callersFunction_template_verifier
Point the Harbor verifier at tmp_path, with stubbed trajectory, security and judges.
tests/tier3/test_metric_applicability_producer.py:310
↓ 3 callersFunction_terminal_console
()
tests/reporting/test_rich_markup_escaping.py:80
↓ 3 callersFunction_text
(value: object)
src/skillevaluator/skillspector_diagnostics.py:15
↓ 3 callersFunction_text
(value, where)
src/skillevaluator/tier3/grader_validation.py:73
↓ 3 callersFunction_text
(value, field: str, *, limit: int = 4096, empty: bool = False)
ci/tier_manifest.py:46
↓ 3 callersFunction_text
(element)
tests/reporting/test_report_version_placement.py:24
↓ 3 callersFunction_text_contains_agent_runtime_failure
(text: str)
src/skillevaluator/tier3/harbor/collector.py:720
↓ 3 callersFunction_tier3_comparison_block_reason
Keep cached comparison evidence from restoring an explicitly unavailable skill arm.
src/skillevaluator/reporting/html.py:566
↓ 3 callersFunction_tier3_evidence
(benchmark: str)
tests/reporting/test_tier3_incomplete_arms.py:133
↓ 3 callersFunction_timed
(call)
tests/tier3/test_dataset_baseline_fairness.py:2075
↓ 3 callersFunction_timeout_recovered_by
Recognize only Harbor's agent-phase deadline followed by a fresh completion. This is evidence classification, not command replay. Setup, transpor
src/skillevaluator/tier3/harbor/collector.py:6705
↓ 3 callersFunction_to_loose
Replace ``${CLAUDE_PLUGIN_ROOT}`` and ``${CLAUDE_PLUGIN_DATA}`` with the loose directories, in every string.
src/skillevaluator/tier3/loose_parts.py:684
↓ 3 callersFunction_tolerated_line
(agent: str)
tests/tier3/test_baseline_timeouts.py:783
↓ 3 callersFunction_tool
(name: str)
tests/test_transport_rules.py:60
↓ 3 callersFunction_tool_call_shell
The shell a tool call says ran its command, and whether this walk models it. A native call can name its shell (Codex's ``exec_command`` keeps
src/skillevaluator/tier3/eval_core/checks.py:7131
↓ 3 callersFunction_tool_calls
Return the step's tool calls (``[]`` when it has none), or ``None`` if malformed.
src/skillevaluator/tier3/harbor/trajectory_timing.py:96
↓ 3 callersFunction_tool_file_path
(args: dict[str, Any])
src/skillevaluator/tier3/eval_core/atif_helpers.py:394
↓ 3 callersFunction_tool_mentions_any_skill
( tool_call: dict[str, Any], expected_skill: str, acceptable_skills: Any = None, )
src/skillevaluator/tier3/eval_core/checks.py:5346
↓ 3 callersFunction_traced_peak_bytes
Return the most memory *callback* held at once, as tracemalloc sees it.
tests/utils/test_structured_data.py:159
↓ 3 callersFunction_trajectory_agent
(trajectory: Mapping[str, Any])
src/skillevaluator/tier3/eval_core/plugin_signals.py:2919
↓ 3 callersFunction_trajectory_cwd
The task's working directory: the agent's recorded cwd, else the first step's.
src/skillevaluator/tier3/eval_core/plugin_signals.py:3264
↓ 3 callersFunction_trajectory_reference_cache_key
(agent_dir: Path, reference_key: str)
src/skillevaluator/tier3/harbor/collector.py:2095
↓ 3 callersFunction_trajectory_skill_invoked
Derive target invocation from executed root and referenced subagent paths.
src/skillevaluator/tier3/harbor/collector.py:6203
← previousnext →3,001–3,100 of 24,741, ranked by callers