MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 1 callersFunctionlaunch_dashboard
Run Streamlit on loopback with explicit paths and usage telemetry disabled.
src/skillevaluator/reporting/dashboard.py:24
↓ 1 callersMethodlaunch_env
(self)
src/skillevaluator/tier3/plugin_native.py:1080
↓ 1 callersFunctionli
(t,e,i)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctionlifecycle
()
tests/reporting/test_terminal_polish.py:583
↓ 1 callersFunctionlift_by_case_type
Split a paired lift into ``cross_component`` cases and all ``other`` cases. Report-only. Each side reuses the paired per-case scores and the pair
src/skillevaluator/tier3/harbor/stats.py:729
↓ 1 callersFunctionlift_ladder
Split the plugin lift into four steps that add up, on one common basis. Report-only. ``arms`` is keyed like :data:`LADDER_ARMS`; an arm that is
src/skillevaluator/tier3/harbor/stats.py:1092
↓ 1 callersFunctionlift_precision
Classify how informative a lift interval is, by case count and width.
src/skillevaluator/tier3/harbor/stats.py:489
↓ 1 callersFunctionline
(row: TokenUsageRow)
src/skillevaluator/reporting/tier3_token_usage.py:278
↓ 1 callersFunctionline_end
(start: int, end: int)
src/skillevaluator/tier3/eval_core/checks.py:975
↓ 1 callersMethodlist_files
Securely list regular files below a contained directory (raises on links). Nothing is pruned: a client loads a nested component folder whatev
src/skillevaluator/plugin_paths.py:218
↓ 1 callersMethodlist_foundation_models
(self)
tests/test_harbor_runtime_preflight.py:1414
↓ 1 callersMethodlist_namespaced_pod
(self, **kwargs: object)
tests/test_harbor_local_mode.py:5827
↓ 1 callersFunctionlisted_description
What the profile says this harness puts in the model's skill list for the description.
src/skillevaluator/harness/overlay.py:742
↓ 1 callersFunctionlisting_budget_chars
Characters the harness reserves for the skill list: a share of the context window when the profile gives ``chars_per_context_token`` and the windo
src/skillevaluator/harness/overlay.py:867
↓ 1 callersFunctionliteral_kwarg_accesses
(node: ast.AST, *, name: str | None = None, attribute: str | None = None)
tests/test_oss_packaging.py:221
↓ 1 callersFunctionlo
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodload
(cls, path: Path, *, client, max_scalar_comparisons=SIMILARITY_DEFAULT_MAX_SCALAR_COMPARISONS)
src/skillevaluator/deduplication/plugin/catalog.py:66
↓ 1 callersFunctionload_baseline_rows
Load the retained without-skill arm.
src/skillevaluator/tier3/baseline_store.py:202
↓ 1 callersMethodload_cache
Deprecated compatibility alias for :meth:`load_catalog`.
src/skillevaluator/embedding/registry.py:713
↓ 1 callersFunctionload_case_marks
Final attempt marks read from a finished run's ``result.json``.
src/skillevaluator/tier3/case_summary.py:1586
↓ 1 callersFunctionload_custom_backend
Load the custom backend behind ``env_mode``; return None for built-in or unknown modes.
src/skillevaluator/tier3_environments.py:388
↓ 1 callersFunctionload_evidence_summary
Per arm and piece, how many scored trials loaded it, and the cases dropped for a load failure. ``pieces[arm][piece]`` counts ``loaded``, ``not_lo
src/skillevaluator/tier3/harbor/loose_parts_arm.py:198
↓ 1 callersFunctionload_expected
Return the expected results for the evaluator that runs the demo. The top-level values are what the release pinned in the starter guide reports.
scripts/demo_starter.py:69
↓ 1 callersFunctionload_kube_config
(**kwargs: str)
tests/test_harbor_local_mode.py:5809
↓ 1 callersFunctionload_license_config
Load license compliance configuration from YAML config. Returns: dict containing: - allowed_licenses: List of SPDX identifiers fo
src/skillevaluator/config/__init__.py:20
↓ 1 callersFunctionload_qualification_policy
Read a bounded regular JSON file and reject incomplete/unknown policy criteria.
src/skillevaluator/tier3/grader_policy.py:25
↓ 1 callersFunctionload_release
(path: Path)
src/skillevaluator/ci_onboarding.py:127
↓ 1 callersFunctionload_report_evidence
Load the signed, bounded JSON evidence for one completed scan attempt.
ci/publish_reports.py:940
↓ 1 callersFunctionload_security_baselines
Return immutable baseline bytes keyed by exact skill path and audit facts.
ci/security_baselines.py:88
↓ 1 callersFunctionload_suite
Strictly validate a suite and re-verify its executable labels.
src/skillevaluator/tier3/grader_validation.py:88
↓ 1 callersFunctionload_unicode_smuggle_patterns
Load Unicode smuggling detection patterns from YAML config.
src/skillevaluator/config/__init__.py:38
↓ 1 callersFunctionlocal_metadata_http_server
Run an ephemeral local HTTP server returning configurable HTTP status codes for probe script tests.
tests/test_harbor_gke_environment.py:52
↓ 1 callersFunctionlocal_repo_slug
Return the ``<group>/<repo>`` slug of ``clone_root``'s git origin, or ``None``. Fails closed: the slug is trusted only when ``clone_root`` is its
src/skillevaluator/plugin_dependencies.py:248
↓ 1 callersFunctionlocal_runtime_error_message
Return a concise, actionable install hint for a missing local runtime.
src/skillevaluator/tier3/harbor/local_runtime.py:560
↓ 1 callersFunctionlocal_runtime_install_command
Return the vendor install hint(s) for the given agent(s).
src/skillevaluator/tier3/harbor/local_runtime.py:554
↓ 1 callersFunctionlog_result
Explain verdicts promptly without replaying arbitrary subprocess logs.
ci/run_tier1.py:289
↓ 1 callersMethodlookalikes
Other skills here whose description has nearly the same words as this one's.
src/skillevaluator/harness/overlay.py:611
↓ 1 callersFunctionlookup
()
src/skillevaluator/validators/endpoint_resolution.py:73
↓ 1 callersFunctionloose_parts_launch_env
The plugin-root variables the loose launch exports, so scripts reading them find the loose root.
src/skillevaluator/tier3/loose_parts.py:257
↓ 1 callersMethodlsp
(self)
src/skillevaluator/plugin_discovery.py:1407
↓ 1 callersFunctionluminance
(color: str)
tests/reporting/test_report_redesign.py:523
↓ 1 callersFunctionma
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctionmain
(argv: Sequence[str] | None = None)
src/skillevaluator/tier3/generate_dataset.py:1830
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/reference_skills/calculator/scripts/calc.py:91
↓ 1 callersFunctionmain
(path: str)
src/skillevaluator/tier3/reference_skills/create-custom-grader/evals/files/create-custom-grader-002/csv-summary/scripts/summarize.py:19
↓ 1 callersFunctionmain
(path: str)
src/skillevaluator/tier3/reference_skills/create-custom-grader/evals/files/create-custom-grader-001/benchmark/code/totals.py:9
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/reference_skills/api-caller/scripts/call_api.py:158
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/reference_skills/api-caller/scripts/parse_openapi.py:120
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/reference_skills/text-analyzer/scripts/analyze_text.py:58
↓ 1 callersFunctionmain
CLI entry point for testing (file-backed).
src/skillevaluator/tier3/reference_skills/task-list/scripts/tasks.py:440
↓ 1 callersFunctionmain
Run the loopback bridge from a trial container.
src/skillevaluator/tier3/harbor/nvidia_build_bridge.py:992
↓ 1 callersFunctionmain
(input_path: Path, output_path: Path)
src/skillevaluator/tier3/harbor/templates/metric.py:84
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/harbor/templates/eval.py:4470
↓ 1 callersFunctionmain
()
src/skillevaluator/tier3/harbor/templates/custom_grader_runner.py:395
↓ 1 callersFunctionmain
()
scripts/script_invocation_differential.py:1364
↓ 1 callersFunctionmain
()
scripts/benchmark_tier2_scale.py:204
↓ 1 callersFunctionmain
(argv: list[str] | None = None)
scripts/check_oss_boundary.py:1064
↓ 1 callersFunctionmain
()
scripts/demo_starter.py:163
↓ 1 callersFunctionmain
()
scripts/ci/smoke_harbor_verifier.py:137
↓ 1 callersFunctionmain
(argv: list[str] | None = None)
scripts/ci/smoke_openclaw_plugins.py:178
↓ 1 callersFunctionmain
(argv: list[str] | None = None)
scripts/ci/check_public_benchmarks.py:2115
↓ 1 callersFunctionmain
()
ci/prepare_security_baselines.py:78
↓ 1 callersFunctionmain
(argv=None)
ci/benchmark_publish.py:373
↓ 1 callersFunctionmain
(argv: list[str] | None = None)
ci/run_tiers.py:800
↓ 1 callersFunctionmain
()
ci/bootstrap_tiers.py:33
↓ 1 callersFunctionmain
()
ci/prepare_comments.py:73
↓ 1 callersFunctionmain
Print ``true`` when the paid jobs must run, or ``false`` when no skill changed.
ci/select_changed_skills.py:348
↓ 1 callersFunctionmain
()
ci/release_image.py:78
↓ 1 callersFunctionmain
()
ci/publish_reports.py:1084
↓ 1 callersFunctionmain
()
ci/render_templates.py:123
↓ 1 callersFunctionmain
()
ci/request_gitlab_feedback.py:104
↓ 1 callersFunctionmain
Run the immutable image or its installed entrypoint; never execute candidate code.
ci/run_tier1_image.py:144
↓ 1 callersFunctionmain
()
ci/smoke_image.py:447
↓ 1 callersFunctionmain
()
ci/eligible_gitlab.py:53
↓ 1 callersFunctionmain
()
ci/run_tier1.py:869
↓ 1 callersFunctionmain
(argv=None)
ci/feedback.py:1135
↓ 1 callersFunctionmain
()
ci/comment_report.py:229
↓ 1 callersFunctionmain
()
tests/fixtures/plugins/bundled-script-hooks/hooks/session-start.py:17
↓ 1 callersFunctionmain
()
tests/fixtures/skills/simple/scripts/summarize.py:21
↓ 1 callersFunctionmain
()
tests/golden/generate_cli_surface.py:27
↓ 1 callersFunctionmake_judge_stub
(captured_prompts: list[str], count: int)
tests/tier3/test_behavior_evidence.py:1220
↓ 1 callersFunctionmake_timestamped_basename
Return ``<prefix>-YYYYMMDDHHMMSS<suffix>`` for report artifacts. Used so each combined ``validate`` run writes a distinct, sortable report fi
src/skillevaluator/utils/helpers.py:17
↓ 1 callersMethodmakefile
(self, mode: str)
src/skillevaluator/model_catalog.py:151
↓ 1 callersFunctionmanaged_install_argv
(agent)
tests/ci/test_all_tier_preflight.py:104
↓ 1 callersFunctionmanifest
()
tests/ci/test_benchmark_report_links.py:9
↓ 1 callersFunctionmanifest_syntax
Every supported format is JSON except the bundle-reference YAML.
src/skillevaluator/plugin_formats.py:328
↓ 1 callersFunctionmark_incomplete_plugin_package
Mark the result incomplete when the prepared plugin package was incomplete. The package's incomplete reasons, such as skill or rule references th
src/skillevaluator/reporting/plugin_reports.py:21
↓ 1 callersFunctionmark_no_skill_script_checks
( checks: object, results: list[dict[str, Any]], *, has_skill: object, skill_metrics_on: b
src/skillevaluator/tier3/eval_core/declarative_checks.py:265
↓ 1 callersMethodmarkdown_components
(self, component_type: str, field_name: str, default_dir: str | None)
src/skillevaluator/plugin_discovery.py:954
↓ 1 callersFunctionmatch_declared_mcp_server
The one declared server that ``observed`` (a server name taken from a tool name) refers to. An exact (case-insensitive) match wins. Otherwise a h
src/skillevaluator/tier3/eval_core/plugin_signals.py:2232
↓ 1 callersMethodmcp
(self)
src/skillevaluator/plugin_discovery.py:1347
↓ 1 callersFunctionmcp_client_for_manifest
The client whose MCP config rules apply to a manifest format (Claude Code by default).
src/skillevaluator/validators/mcp_static.py:170
↓ 1 callersFunctionmcp_input_schemas
``{server: {tool: schema}}`` from a proof's host probes, at most ``budget`` bytes of compact JSON in all. Each kept tool is counted as written: i
src/skillevaluator/tier3/mcp_proof.py:967
↓ 1 callersFunctionmcp_pinning_summary
C1 ``mcp.pinning`` / C2 ``mcp_pinning`` of ``declarations`` (see :func:`summarize_pinning`).
src/skillevaluator/plugin_mcp.py:175
↓ 1 callersMethodmcp_summary
(self)
src/skillevaluator/plugin_inventory.py:170
↓ 1 callersFunctionmeasure
(trials: list[dict[str, Any]], field: str, complete: bool)
src/skillevaluator/tier3/harbor/resource_usage.py:87
↓ 1 callersFunctionmeasure
(stages: dict[str, Any], name: str, action)
scripts/benchmark_tier2_scale.py:80
↓ 1 callersFunctionmeasured_condition_usage
Sum observed trajectory counters; missing measurements remain unavailable.
src/skillevaluator/tier3/baseline_store.py:54
↓ 1 callersFunctionmember_overlap
(left, right)
src/skillevaluator/deduplication/plugin/profile.py:66
↓ 1 callersMethodmerge_advisories_with_prefix
Carry a *passing* sub-result's nonblocking findings into this one. ``merge_with_prefix`` is for a sub-result that failed. A member can
src/skillevaluator/models/result.py:817
← previousnext →8,801–8,900 of 24,741, ranked by callers