MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 1 callersFunction_normalize_harbor_viewer
Normalize one source before choosing primary or legacy summary evidence.
src/skillevaluator/reporting/harbor_viewer.py:43
↓ 1 callersFunction_normalize_harbor_viewer_metadata
(raw: object)
src/skillevaluator/evaluation/tier3_report.py:3882
↓ 1 callersFunction_normalize_heading
(value: object, source: str, location: str)
src/skillevaluator/quality_config.py:137
↓ 1 callersFunction_normalize_openclaw_server
Map one native OpenClaw server onto the shared fields (see :func:`_normalize_dialect`).
src/skillevaluator/plugin_mcp.py:692
↓ 1 callersFunction_normalize_opencode_tool
Map OpenCode tool names and inputs to ATIF tool-call fields.
src/skillevaluator/tier3/eval_core/log_converters.py:405
↓ 1 callersFunction_normalize_required_judge_result
(metric, result, *, allow_not_applicable=False)
src/skillevaluator/tier3/harbor/templates/eval.py:3934
↓ 1 callersFunction_normalize_sensitive_path_text
Lowercase *text*, normalize its path words, and rewrite home directories to ``~``.
src/skillevaluator/tier3/eval_core/checks.py:926
↓ 1 callersFunction_normalize_skill_names
Normalize dataset-provided skill name fields into a de-duplicated list.
src/skillevaluator/tier3/eval_core/checks.py:3183
↓ 1 callersMethod_normalize_usernames
Lower-case, strip, and drop empties from an iterable of usernames.
src/skillevaluator/validators/security.py:3065
↓ 1 callersFunction_normalized_metric_name_parts
(name: str)
src/skillevaluator/tier3/harbor/templates/custom_grader_runner.py:171
↓ 1 callersFunction_normalized_relative_path
Return the platform-independent NFC path used by the recipe.
src/skillevaluator/publication_identity.py:582
↓ 1 callersFunction_normalized_text
(text: object)
src/skillevaluator/evaluation/insights_judge.py:494
↓ 1 callersFunction_not_a_runner
The classification of a launch that runs no package runner.
src/skillevaluator/validators/mcp_static.py:2567
↓ 1 callersMethod_not_applicable
(result: ValidationResult, parity: dict[str, Any], reason: str)
src/skillevaluator/validators/claude_plugin_validate.py:557
↓ 1 callersFunction_not_applicable_dimensions
Dimensions left unscored because every source evaluator was not applicable.
src/skillevaluator/evaluation/tier3_report.py:2107
↓ 1 callersFunction_not_applicable_evaluators
Evaluators with no score in the reported arm because every trial recorded them as N/A.
src/skillevaluator/evaluation/tier3_report.py:2084
↓ 1 callersFunction_not_applicable_result
(check_id: str)
src/skillevaluator/tier1/commands.py:401
↓ 1 callersFunction_not_globally_reachable
Any other special-purpose address (IANA: not globally reachable, multicast, or reserved) is not public.
src/skillevaluator/validators/mcp_static.py:2888
↓ 1 callersMethod_not_run
(agent: str, arm: str, case_id: str, attempt: int)
src/skillevaluator/tier3/harbor/run_budget.py:401
↓ 1 callersMethod_notifyStateChanges
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunction_npm_call_value
The value of npm's ``-c``/``--call`` among a runner's options (npm exec reads them anywhere before ``--``).
src/skillevaluator/validators/mcp_static.py:660
↓ 1 callersFunction_npm_invocation
An npm package runner: every ``-p``/``--package`` value, else the first positional argument. The runner's options end at the package it runs, or,
src/skillevaluator/validators/mcp_static.py:2167
↓ 1 callersFunction_npm_report
(data: dict[str, Any], pins: dict[str, str], source: str, exit_code: int)
src/skillevaluator/plugin_dependency_audit.py:514
↓ 1 callersFunction_npm_source
(inputs: AuditInputs, source: str, data: Any, *, lockfile: bool)
src/skillevaluator/plugin_dependency_audit.py:208
↓ 1 callersFunction_number
(value: Any, minimum: float = 0, maximum: float = 1)
src/skillevaluator/tier3/history.py:79
↓ 1 callersFunction_number
(value: object, *, minimum: float)
ci/summary_metrics.py:33
↓ 1 callersFunction_number_may_fail
Whether a numeric attribute may fail on ``value``: as ``_integer_may_fail``, where a decimal, ``0x10`` and ``16#ff`` are numbers too and nothing (
src/skillevaluator/tier3/eval_core/script_invocation.py:738
↓ 1 callersFunction_numbers
(values: list[str])
src/skillevaluator/tier3/reference_skills/create-custom-grader/evals/files/create-custom-grader-002/csv-summary/scripts/summarize.py:9
↓ 1 callersFunction_numeric_reward_payload
(result, overall)
src/skillevaluator/tier3/harbor/templates/eval.py:4156
↓ 1 callersFunction_observations
Each result of ``step``. ``flagged`` is the structured outcome of :func:`_result_flagged`. Harbor puts the flag of an orphan Claude Code resu
src/skillevaluator/tier3/eval_core/plugin_signals.py:1952
↓ 1 callersFunction_observe_private_trials
Hand existing observations to an active private hook without affecting ordinary collection.
src/skillevaluator/tier3/harbor/collector.py:7569
↓ 1 callersFunction_occurrence_record_count
(issue: object)
src/skillevaluator/validators/skillspector_scope.py:475
↓ 1 callersFunction_only_batched
Whether ``refs`` were used, but never by a call that opened a window (one read of several SKILL.md files).
src/skillevaluator/tier3/eval_core/plugin_signals.py:4506
↓ 1 callersFunction_only_definitions
(body: Sequence[ast.stmt], local_functions: frozenset[str])
src/skillevaluator/tier3/generate_dataset.py:258
↓ 1 callersFunction_open_directory_no_follow
(path: Path)
src/skillevaluator/publication_identity.py:292
↓ 1 callersFunction_open_manifest_root
(manifest: _TreeManifest)
src/skillevaluator/tier3/harbor/secure_copy.py:712
↓ 1 callersMethod_open_posix
(self, parts: tuple[str, ...])
src/skillevaluator/deduplication/utils/skill_collector.py:199
↓ 1 callersFunction_open_posix_child_directory
Open the frame's next admitted child relative to its held descriptor, still unchanged.
src/skillevaluator/utils/secure_fs.py:1250
↓ 1 callersFunction_open_posix_parent
Open the report parent component-by-component without following links.
src/skillevaluator/reporting/base.py:181
↓ 1 callersMethod_open_windows
(self, parts: tuple[str, ...])
src/skillevaluator/deduplication/utils/skill_collector.py:245
↓ 1 callersFunction_open_windows_child_directory
Open the frame's next admitted child relative to its pinned handle, still unchanged.
src/skillevaluator/utils/secure_fs.py:1407
↓ 1 callersFunction_openclaw_built_entries
Bounded built-runtime candidates for one TypeScript entry, in OpenClaw's lookup order.
src/skillevaluator/plugin_discovery.py:175
↓ 1 callersMethod_openclaw_entry
(self, field_name: str, raw: Any, *, allow_missing: bool = False)
src/skillevaluator/plugin_discovery.py:1264
↓ 1 callersFunction_openclaw_gates
(data: dict[str, Any])
src/skillevaluator/harness/overlay.py:389
↓ 1 callersFunction_openclaw_read_chars
SKILL.md size as OpenClaw's tool-result cap measures it.
src/skillevaluator/harness/overlay.py:218
↓ 1 callersFunction_opencode_call_errors
OpenCode tool calls that failed, read from the raw ``opencode.txt``: ``call id -> error text``. OpenCode's JSON stream ends each tool part with `
src/skillevaluator/tier3/plugin_trace_artifacts.py:337
↓ 1 callersFunction_opencode_exit_code
Return OpenCode bash ``metadata.exit`` when present.
src/skillevaluator/tier3/eval_core/log_converters.py:462
↓ 1 callersFunction_opencode_markdown
(text: NativeTextComponent, permission: Mapping[str, Any] | None = None)
src/skillevaluator/tier3/plugin_native.py:1962
↓ 1 callersFunction_opencode_output_payload
(state: dict[str, Any])
src/skillevaluator/tier3/eval_core/log_converters.py:429
↓ 1 callersFunction_opencode_permission
Translate a Claude subagent's ``tools``/``disallowedTools`` into an OpenCode ``permission`` block. Returns ``(permission, narrowed, unenforceable
src/skillevaluator/tier3/plugin_native.py:1909
↓ 1 callersFunction_opencode_tool_observation
Build a bounded observation that retains status, exit code, and errors.
src/skillevaluator/tier3/eval_core/log_converters.py:476
↓ 1 callersFunction_options
(command: click.Command, ctx: click.Context)
src/skillevaluator/cli_help.py:182
↓ 1 callersMethod_other_main_environment_values
( self, retained_name: str, retained_value: str, )
src/skillevaluator/tier3/harbor/secure_docker_environment.py:2262
↓ 1 callersFunction_outcome
``(observation text, succeeded)`` for one call (see the module docs). A structured outcome wins: any failure flag fails the call, and the har
src/skillevaluator/tier3/eval_core/plugin_signals.py:2056
↓ 1 callersFunction_outcome_rows
( run_dir: Path, agent: str, condition: str, metric: str, *, custom_only: bool )
src/skillevaluator/tier3/interactions.py:88
↓ 1 callersMethod_output_secret_values
(self, env: dict[str, str] | None, exec_env: dict[str, str])
src/skillevaluator/tier3/harbor/local_environment.py:1833
↓ 1 callersFunction_overall_aligned
A case scored only by its overall on one side compares with the other side's overall (dimension mean).
src/skillevaluator/tier3/harbor/stats.py:415
↓ 1 callersFunction_overall_status
( results: list[ValidationResult], ae: dict[str, Any] | None, benchmark_policy: dict[str, bool],
src/skillevaluator/reporting/benchmark.py:1185
↓ 1 callersFunction_overall_verdict_fields
Return canonical and visibly callout-shaped verdict fields.
scripts/ci/check_public_benchmarks.py:1169
↓ 1 callersFunction_overview_ids
(payload: dict[str, Any])
tests/reporting/test_tier3_failure_reasons.py:860
↓ 1 callersFunction_owned_comments
(api, repository: str, number: int, bot: str)
ci/comment_report.py:124
↓ 1 callersFunction_p02_trajectory
(*, codex: bool)
tests/tier3/test_f90_mcp_failure_text.py:86
↓ 1 callersFunction_pack_paragraphs
Pack consecutive paragraphs of an oversized section up to the embedding cap. Paragraphs shorter than ``min_chars`` stay attached to a neighbour i
src/skillevaluator/deduplication/utils/chunker.py:189
↓ 1 callersFunction_package_runs
Classify every package launch, retaining bounded samples of the strongest risks. Commands are already byte-bounded. Each stage is parsed whole wi
src/skillevaluator/plugin_component_risk.py:2615
↓ 1 callersFunction_packaged_uv
Return the uv executable installed by SkillEvaluator's ``uv`` dependency. A ``pip install --user`` or ``--target`` layout puts it in a scripts
src/skillevaluator/utils/tool_runner.py:88
↓ 1 callersFunction_pair_completeness
The plugin arm against one control arm: missing cases, failed arms and attempt shortfall.
src/skillevaluator/tier3/harbor/stats.py:1515
↓ 1 callersFunction_paired_summary
(contract: ExternalMetrics, name: str)
src/skillevaluator/tier3/external_metrics.py:212
↓ 1 callersFunction_param_strings
Return the strings that *test* is parametrized with for *argname*.
tests/tier3/test_security_finding_evidence.py:4390
↓ 1 callersFunction_param_to_dict
(param: click.Parameter)
tests/golden/_cli_surface.py:31
↓ 1 callersFunction_parse_args
Parse raw args into (operation, task_id, content, error).
src/skillevaluator/tier3/reference_skills/task-list/scripts/tasks.py:360
↓ 1 callersFunction_parse_args
(argv: list[str] | None)
scripts/check_oss_boundary.py:1056
↓ 1 callersFunction_parse_argument_rule
(raw: Any, rule_where: str, errors: _FieldErrors)
src/skillevaluator/tier3/eval_core/plugin_signals.py:951
↓ 1 callersMethod_parse_bridge_origin
(value: str)
src/skillevaluator/tier3/harbor/local_agents.py:754
↓ 1 callersFunction_parse_claude_marketplace
( data: Any, root: Path, plugin_root: Path, plugin_name: str | None )
src/skillevaluator/plugin_manifest.py:633
↓ 1 callersFunction_parse_config_text
Parse one bounded YAML document whose mappings, merge sources included, repeat no key. A YAML error never repeats the source line. A marked error
src/skillevaluator/tier3/evals_config.py:194
↓ 1 callersFunction_parse_expression_items
Run the SPDX expression grammar over grouped names, operators and parentheses.
src/skillevaluator/validators/license.py:111
↓ 1 callersMethod_parse_frontmatter
(content: str)
src/skillevaluator/validators/quality_score.py:480
↓ 1 callersFunction_parse_json_argument
Parse and type-check a JSON command-line argument.
src/skillevaluator/tier3/reference_skills/api-caller/scripts/call_api.py:142
↓ 1 callersFunction_parse_legacy_ipv4
Parse inet_aton-style IPv4 (decimal/hex/octal parts, 1-4 components).
src/skillevaluator/validators/mcp_static.py:2816
↓ 1 callersFunction_parse_retry_after
Parse a Retry-After header as seconds or HTTP date, falling back to default.
src/skillevaluator/tier3/harbor/templates/eval.py:1732
↓ 1 callersFunction_parse_rule_section
Validate one rule section. With a base, keys the section omits keep the base value.
src/skillevaluator/quality_config.py:249
↓ 1 callersFunction_parse_storage_mib
Parse a Kubernetes storage quantity string (Mi/Gi/M/G) into MiB.
src/skillevaluator/tier3/harbor/gke_environment.py:187
↓ 1 callersFunction_parse_toml_dotted_key
Parse a TOML key or table path into unquoted, escape-decoded segment names.
src/skillevaluator/tier3/harbor/adapter.py:4815
↓ 1 callersMethod_parse_truncated_response
Parse kept partial evidence; subclasses can reject unusable salvage.
src/skillevaluator/inference/client.py:899
↓ 1 callersFunction_parse_unit_starts
Where each top-level unit the shell parses before running it starts, and the text with its comments blanked. The first is the start of the te
src/skillevaluator/tier3/eval_core/script_invocation.py:1770
↓ 1 callersFunction_parse_upload_cached
(file_id: str, source: str, name: str, _contents: bytes)
src/skillevaluator/reporting/dashboard_app.py:115
↓ 1 callersMethod_parse_verdicts
(response_text: str)
src/skillevaluator/inference/finding_verifier.py:176
↓ 1 callersFunction_parse_whole_json_object
Return *text* parsed as one JSON object, or None when it holds anything else.
src/skillevaluator/tier3/harbor/templates/eval.py:2585
↓ 1 callersFunction_parser
()
scripts/classify_ci_changes.py:173
↓ 1 callersFunction_parser
()
scripts/select_docs_tests.py:104
↓ 1 callersFunction_parser
()
scripts/ci/check_public_benchmarks.py:2075
↓ 1 callersFunction_partial_agent_eval_result
Normalize an engine run that produced usable results alongside errors. A run where one agent crashed but another scored is still a run: discardin
src/skillevaluator/cli.py:787
↓ 1 callersFunction_partial_known_prefix_length
(text: str)
src/skillevaluator/tier3/harbor/stream_redaction.py:332
↓ 1 callersFunction_partial_load_reason
Why a component is ``not_loaded`` over the trials, with how many trials did load it. The first trial's ``not_loaded`` text alone ("did not load p
src/skillevaluator/tier3/plugin_native.py:2792
↓ 1 callersFunction_partial_pass_summary
(summary: dict[str, Any])
src/skillevaluator/tier3/harbor/collector.py:5098
↓ 1 callersFunction_partition_views
Give each source file to its deepest skill, or to the plugin root. Readers inspect the complete, already verified snapshot once. Partition ro
src/skillevaluator/tier1/plugin_checks.py:158
↓ 1 callersFunction_pass_at_k_cases
Return an arm's pass@k record per case, under the case names the rows use. Two keys can canonicalize onto one case, Harbor's cut of a name beside
src/skillevaluator/tier3/token_usage.py:298
↓ 1 callersFunction_pass_rate_changes
(agents: Mapping[str, Mapping[str, Any]])
src/skillevaluator/tier3/next_steps.py:114
↓ 1 callersFunction_pass_threshold_from_policy
(attempt_policy: dict[str, Any])
src/skillevaluator/evaluation/tier3_report.py:4526
↓ 1 callersFunction_passing_skill_suggestions
Generate improvement suggestions even when all metrics pass.
src/skillevaluator/tier3/harbor/report.py:1025
← previousnext →7,001–7,100 of 24,741, ranked by callers