MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 2 callersFunction_activation_status
(raw: Any)
src/skillevaluator/evaluation/tier3_feedback_chat.py:743
↓ 2 callersFunction_activation_summary
(info: dict[str, Any])
src/skillevaluator/evaluation/tier3_report.py:76
↓ 2 callersFunction_adaptive_unsigned_percent_decimals
Return the existing bounded precision policy for one unsigned percentage.
src/skillevaluator/reporting/html.py:1863
↓ 2 callersMethod_add_apps
One ``app`` component per alias in each ``.app.json`` ``apps`` map; a file without aliases is one.
src/skillevaluator/plugin_discovery.py:1090
↓ 2 callersFunction_add_image
(inputs: AuditInputs, image: str, source: str)
src/skillevaluator/plugin_dependency_audit.py:271
↓ 2 callersMethod_add_license_mismatch_finding
Fail closed when two allowed licenses disagree, without marking allowed.
src/skillevaluator/validators/license.py:565
↓ 2 callersFunction_add_server_tool
(bucket: dict[str, Any], label: str)
src/skillevaluator/tier3/eval_core/plugin_signals.py:4170
↓ 2 callersFunction_admission
( root: Path, *, max_paths: int = 10, excluded_dirs: tuple[str, ...] = () )
tests/utils/test_secure_fs_platform.py:25
↓ 2 callersFunction_advisory
(ecosystem: str, source: str, name: str, version: str, advisory: str, severity: str)
src/skillevaluator/plugin_dependency_audit.py:485
↓ 2 callersFunction_advisory_agent_eval_skip_message
(results: list[ValidationResult])
src/skillevaluator/reporting/benchmark.py:1158
↓ 2 callersMethod_advisory_metadata
Bind an advisory to the scanned version and the exact pin that declared it.
src/skillevaluator/validators/dependencies.py:906
↓ 2 callersMethod_advisory_package_key
Return a PEP 503 name and parsed version, so both scanners' spellings compare equal.
src/skillevaluator/validators/dependencies.py:762
↓ 2 callersFunction_after_copied_context
(first_new_step: dict[str, Any] | None)
tests/tier3/test_trajectory_timing.py:446
↓ 2 callersFunction_agent
(score: float, *, baseline_scored: int = 1, recovered: int = 0)
tests/reporting/test_tier3_completed_agent_docs.py:70
↓ 2 callersFunction_agent_artifact_dirs
Return the agent directories Harbor populates for one trial. Single-step trials keep them at ``<trial>/agent``; multi-step trials keep one pe
src/skillevaluator/tier3/harbor/runtime_preflight.py:983
↓ 2 callersFunction_agent_cell
( with_case: Mapping[str, Any] | None, without_case: Mapping[str, Any] | None, *, comparable:
src/skillevaluator/tier3/case_summary.py:673
↓ 2 callersFunction_agent_credentials
Resolve operator-owned credentials for exactly one agent runtime.
src/skillevaluator/tier3/harbor/runner.py:3152
↓ 2 callersFunction_agent_eval
(volume: dict)
tests/reporting/test_evidence_volume_rendering.py:78
↓ 2 callersFunction_agent_eval_condition_attempt_coverage_complete
Validate the two canonical conditions against one agent aggregate.
src/skillevaluator/reporting/base.py:1074
↓ 2 callersFunction_agent_eval_evaluated_at_datetime
Parse one timezone-aware evaluation instant within bounded clock skew.
src/skillevaluator/reporting/base.py:734
↓ 2 callersFunction_agent_eval_execution_errors_clear
Return whether succeeded Tier 3 evidence contains no execution errors.
src/skillevaluator/reporting/base.py:1124
↓ 2 callersFunction_agent_eval_fingerprint_text
Return a bounded canonical text projection without scanning a huge tail.
src/skillevaluator/reporting/base.py:1541
↓ 2 callersFunction_agent_eval_publication_target_conflict_markers
Return bounded conflict markers from both canonical Tier 3 copies.
src/skillevaluator/reporting/base.py:890
↓ 2 callersFunction_agent_eval_rejected_truth_state
Read only bounded root truth fields from a rejected Tier 3 payload.
src/skillevaluator/reporting/base.py:698
↓ 2 callersFunction_agent_eval_target_skill_issue
Return a fail-closed reason when Tier 3 targets a different skill.
src/skillevaluator/reporting/base.py:2129
↓ 2 callersFunction_agent_generated_output_paths
(agent_dir: Path)
src/skillevaluator/tier3/harbor/collector.py:286
↓ 2 callersFunction_agent_has_dimension_signals
(agent: dict[str, Any], dim_id: str, *, baseline_only: bool = False)
src/skillevaluator/reporting/benchmark.py:1791
↓ 2 callersFunction_agent_identity
(agent: Any)
src/skillevaluator/tier3/harbor/collector.py:2131
↓ 2 callersFunction_agent_import_path
Select the agent wrapper the provider and the execution backend both require.
src/skillevaluator/tier3/harbor/runner.py:3667
↓ 2 callersFunction_agent_log_runtime_failure_reason
Return the most specific agent log/runtime startup failure, if present.
src/skillevaluator/tier3/harbor/collector.py:886
↓ 2 callersFunction_agent_provider_config
Describe the API provider the selected agent will actually call.
src/skillevaluator/tier3/harbor/runner.py:3216
↓ 2 callersFunction_agent_score_headlines
Build the compact verdict, lift, and per-arm check-count summary.
src/skillevaluator/tier3/result_display.py:590
↓ 2 callersFunction_agent_score_policy
Return the with-skill score policy used to rank one agent's raw Harbor data. A recorded policy wins. Artifacts without one use the policy for the
src/skillevaluator/evaluation/tier3_report.py:4714
↓ 2 callersFunction_aggregate_layout
Member names of a full three-tier aggregate as its producers retain them. ``diagnostics`` holds one worker-diagnostics tree (files, folders) reta
tests/ci/test_feedback.py:1173
↓ 2 callersFunction_aggregate_selects_enriched
(env)
tests/ci/test_gateway_benchmark_workflows.py:73
↓ 2 callersMethod_alias_bomb
(levels: int)
tests/validators/test_schema.py:1490
↓ 2 callersFunction_aliases_target
(entry: ContentEntry, target: ContentEntry)
src/skillevaluator/tier3/decoys.py:140
↓ 2 callersFunction_allow_shapes_of
:func:`_allow_shapes` of comment-free ``code`` and its escape-decoded form ``decoded``.
src/skillevaluator/plugin_component_risk.py:1912
↓ 2 callersFunction_allowed_env
Variable names the user opted in for this server (by name, by its URL host, or for every server).
src/skillevaluator/tier3/mcp_proof.py:179
↓ 2 callersFunction_allows_plain_http
Loopback, or a host the operator named for plain-HTTP catalog reads. Resolution is deliberately absent. An entry matches one whole host as wr
src/skillevaluator/model_catalog.py:773
↓ 2 callersMethod_alternation
(self)
src/skillevaluator/plugin_component_risk.py:1268
↓ 2 callersFunction_ambiguous_resource_notice
(source: str)
src/skillevaluator/reporting/dashboard_data.py:292
↓ 2 callersFunction_analyze_allowed_tools
Grant findings for a command's or skill's ``allowed-tools``, which Claude Code pre-approves. For a client in :data:`CLIENTS_IGNORING_ALLOWED_TOOL
src/skillevaluator/plugin_component_risk.py:1004
↓ 2 callersFunction_anthropic_record
()
tests/test_skillspector_provenance.py:127
↓ 2 callersFunction_api_key_assignment
Build a representative hardcoded-key assignment without embedding it in source.
tests/validators/test_security.py:102
↓ 2 callersFunction_api_key_line
Build a representative API-key finding without embedding it in source.
tests/reporting/test_reporters.py:63
↓ 2 callersFunction_append_context_sources
(line: str)
src/skillevaluator/tier3/harbor/adapter.py:2895
↓ 2 callersFunction_append_native_task_name_suffix
Append dual-arm suffix to [task] name in a native task's task.toml.
src/skillevaluator/tier3/harbor/adapter.py:4988
↓ 2 callersFunction_apply_arg
(payload: str, *, stage_scope: bool)
src/skillevaluator/tier3/harbor/adapter.py:2850
↓ 2 callersFunction_apply_binding_builtin
Apply what a builtin that binds variables does, and say whether it was one. ``export f=run.py``, ``readonly f=run.py`` and, where the shell has t
src/skillevaluator/tier3/eval_core/script_invocation.py:1051
↓ 2 callersFunction_apply_env
(payload: str)
src/skillevaluator/tier3/harbor/adapter.py:2872
↓ 2 callersFunction_apply_fallback_open_file_mode
Apply a mode through the strongest API available and verify identity.
src/skillevaluator/tier3/harbor/secure_copy.py:1711
↓ 2 callersFunction_apply_patch_bodies_are_data
Return whether stripped apply_patch bodies are only data, judged from all of the shell text around them. Every apply_patch name in *shell_texts*
src/skillevaluator/tier3/eval_core/checks.py:1591
↓ 2 callersFunction_apply_patch_call
Return ``(patch, workdir, shell)`` for an apply_patch tool call or a shell apply_patch command. ``workdir`` is the directory relative header path
src/skillevaluator/tier3/eval_core/checks.py:1276
↓ 2 callersFunction_apply_patch_command_evidence
Return a shell apply_patch command as evidence: the command before the patch body, secrets masked. The text before the body is masked before it i
src/skillevaluator/tier3/eval_core/checks.py:1499
↓ 2 callersFunction_apply_patch_heredoc
(prefix, header)
tests/tier3/test_private_plugin_security.py:499
↓ 2 callersFunction_apply_patch_heredoc_pieces
Split a shell *command* into text the shell reads and the stripped heredoc bodies it feeds apply_patch. The result alternates shell text and bodi
src/skillevaluator/tier3/eval_core/checks.py:1532
↓ 2 callersFunction_apply_retention_outcome
Apply actual artifact filesystem truth to the returned result.
src/skillevaluator/tier3/harbor/runner.py:7006
↓ 2 callersFunction_argv
The argv a runnable MCP declaration runs, through its launch wrappers, or ``None`` when it has no command. A wrapper that names no command (``env
src/skillevaluator/validators/mcp_static.py:1977
↓ 2 callersFunction_argv_credential_spans
Return the text the argv rule masks in *text* and what that hides from later rules. *flag_values* are the flag values the flag rule masks in *tex
src/skillevaluator/tier3/eval_core/checks.py:4659
↓ 2 callersFunction_arithmetic_names
The names the arithmetic in a command assigns (``_arithmetic_texts``).
src/skillevaluator/tier3/eval_core/script_invocation.py:2288
↓ 2 callersFunction_arm
(totals, *, status="complete", measured=2, reasons=None)
tests/reporting/test_telemetry_view.py:26
↓ 2 callersFunction_arm_failures
An agent that failed the named arms; a failed baseline alone leaves the with-skill arm scored.
tests/test_tier3_result_display.py:311
↓ 2 callersFunction_artifact_fingerprint
Return identity and content-change fields from one stat API family.
src/skillevaluator/tier3/output_provenance.py:43
↓ 2 callersFunction_as_float
(value: Any)
src/skillevaluator/evaluation/tier3_report.py:5173
↓ 2 callersFunction_as_nonnegative_int
(value: Any)
src/skillevaluator/evaluation/tier3_report.py:5177
↓ 2 callersFunction_ascii_dashes
(text: str)
tests/test_metrics_reference_docs.py:79
↓ 2 callersFunction_assert_html_verdict_is_incomplete
The primary result and the evidence summary say Incomplete, without a passing claim.
tests/reporting/test_incomplete_scans.py:28
↓ 2 callersFunction_assert_manifest_check_failed
(results: list[ValidationResult])
tests/test_artifact_applicability.py:218
↓ 2 callersFunction_assert_one_uv_version
(pins: Sequence[tuple[str, str | None]])
tests/test_ci_workflows.py:530
↓ 2 callersFunction_assert_readable_text
(locator, *state)
tests/reporting/test_report_browser.py:448
↓ 2 callersMethod_assert_refused
(result: ValidationResult, version: str)
tests/validators/test_security.py:9932
↓ 2 callersFunction_assert_runtime_projection
(task: Path, target: Path, workspace: Path)
tests/test_harbor_runtime_skill_isolation.py:141
↓ 2 callersFunction_assert_safe_readiness_check
(dockerfile: Path)
tests/test_harbor_compose_readiness.py:15
↓ 2 callersFunction_assert_sensitive_repo_files_are_excluded
(repo: Path, skill: Path, env_dir: Path)
tests/test_harbor_adapter_safety.py:124
↓ 2 callersFunction_assert_terminal_summary_lists_a_failure
The Tier 1 terminal summary counts the check as failed and points at a gate failure. A check that is not counted as a confirmed failure is listed
tests/validators/test_skillspector_hook_surface.py:395
↓ 2 callersFunction_assessment_status
Project one assessment onto the card's uppercase verdict vocabulary.
src/skillevaluator/reporting/benchmark.py:1199
↓ 2 callersFunction_assignment_name_is_kept
Return whether the assignment rule keeps the value after the name *match*. So it does for a name that names no credential, a location, type or
src/skillevaluator/tier3/eval_core/checks.py:5099
↓ 2 callersFunction_assignment_name_matches
Return, in order, each name that may start an assignment in *text*, and where its sed expression ends. That is the end of the "s/PATTERN/REPLACEM
src/skillevaluator/tier3/eval_core/checks.py:5204
↓ 2 callersFunction_assignment_value_reading
Return how the assignment rule reads the value after the name *match*: (pattern, start, header value prefix). The value is *pattern* matched at *
src/skillevaluator/tier3/eval_core/checks.py:5132
↓ 2 callersFunction_attach_harbor_evidence_to_recommendations
( recommendations: list[dict[str, Any]], evidence_links: list[dict[str, Any]], *, start: int =
src/skillevaluator/evaluation/tier3_report.py:4120
↓ 2 callersFunction_attach_harbor_evidence_to_suggestions_v2
( suggestions: list[dict[str, Any]], evidence_links: list[dict[str, Any]], )
src/skillevaluator/evaluation/tier3_report.py:4170
↓ 2 callersFunction_attached_argv_password_spans
Return where the value of each attached "-pVALUE" element after a login or MySQL client element is. The login or MySQL client element must lie in
src/skillevaluator/tier3/eval_core/checks.py:4710
↓ 2 callersFunction_attempt
(root, ordinal, *, timeout)
tests/tier3/test_private_recovered_timeout_review2.py:17
↓ 2 callersMethod_attempt_key
(attempt: dict[str, Any])
src/skillevaluator/tier3/harbor/run_budget.py:253
↓ 2 callersFunction_attempt_sort_key
Sort attempts by explicit attempt label, then Harbor start time.
src/skillevaluator/tier3/harbor/collector.py:4684
↓ 2 callersFunction_attempt_usage
Return an attempt's token fields as a report trial carries them. ``tokens`` holds the valid counters, and ``invalid_token_counters`` names th
src/skillevaluator/tier3/harbor/report_data.py:739
↓ 2 callersFunction_authenticated_generated_output_latest_alias
Recognize only ``latest`` links to an authenticated immediate run child.
src/skillevaluator/tier3/harbor/adapter.py:644
↓ 2 callersMethod_author_identity
Return the skill's declared author identity, if one is parseable.
src/skillevaluator/validators/security.py:3128
↓ 2 callersFunction_authorization_credential
Whether a command word writes an ``Authorization`` header with a literal credential.
src/skillevaluator/validators/mcp_static.py:263
↓ 2 callersFunction_authorization_value_credential
Whether an ``Authorization`` value carries a literal credential after its scheme (``Token abc``, ``Bot x``). Only the text the plugin ships count
src/skillevaluator/validators/mcp_static.py:252
↓ 2 callersFunction_average_multistep_custom_metrics
Mirror Harbor 0.22 MEAN aggregation by zero-filling missing custom keys.
src/skillevaluator/tier3/harbor/collector.py:4043
↓ 2 callersFunction_backticked_identity_present
(value: str)
scripts/ci/check_public_benchmarks.py:1861
↓ 2 callersFunction_bare_host
Lower-case host without brackets, an IPv6 zone ID, or a trailing dot (nothing decoded).
src/skillevaluator/validators/mcp_static.py:2791
↓ 2 callersFunction_bare_words
Command words without shell punctuation attached to them: ``(cd`` is ``cd``, and ``./x.sh)`` and ``"${CLAUDE_PLUGIN_ROOT}/x.sh";`` (quotes already
src/skillevaluator/plugin_component_risk.py:1790
↓ 2 callersFunction_baseline_alias_source_key
( target_skill: Path, reference_skills_dir: Path | None, workspace_skill_paths: Sequence[Path] | N
src/skillevaluator/tier3/harbor/adapter.py:1009
↓ 2 callersFunction_baseline_alias_validation_matches
( validation: _BaselineAliasValidation, target_skill: Path, reference_skills_dir: Path | None,
src/skillevaluator/tier3/harbor/adapter.py:1050
↓ 2 callersFunction_bash
(command)
tests/tier3/test_shared_check_adapters.py:161
↓ 2 callersFunction_bash_grant_label
(entry: str)
src/skillevaluator/plugin_component_risk.py:607
↓ 2 callersFunction_bedrock_credential_process
(session: object | None)
src/skillevaluator/tier3/harbor/runtime_preflight.py:227
← previousnext →3,501–3,600 of 24,741, ranked by callers