MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 1 callersFunction_simple_assignments
Simple ``NAME=value`` assignments in ``text``, each value expanded with the earlier ones (bounded).
src/skillevaluator/plugin_component_risk.py:2470
↓ 1 callersFunction_simulate
(n_cases: int, sims: int, seed: int)
tests/tier3/test_private_lift_small_n.py:30
↓ 1 callersFunction_sized
A skill whose ``fact`` (a body-length measure) is exactly ``target``.
tests/harness/test_overlay.py:754
↓ 1 callersMethod_skill
(self, tmp_path: Path)
tests/test_benchmark_source_provenance.py:1139
↓ 1 callersFunction_skill_checks
()
src/skillevaluator/tier1/plugin_checks.py:69
↓ 1 callersFunction_skill_description
(skill_path: Path | None)
src/skillevaluator/tier3/case_summary.py:1354
↓ 1 callersFunction_skill_domain_tokens
(skill: dict[str, Any])
src/skillevaluator/tier3/generate_dataset.py:365
↓ 1 callersFunction_skill_files_re
Wording that names the skill's own files: skills/<name>, <name>/SKILL.md, or reading a bare SKILL.md.
src/skillevaluator/tier3/evals_spec.py:717
↓ 1 callersFunction_skill_identity_names
Directory and declared identities, with bounded frontmatter input/parsing.
src/skillevaluator/tier3/evals_spec.py:1032
↓ 1 callersFunction_skill_manifest_text
A member skill's ``SKILL.md`` (bounded, no-follow read), or ``""`` when it has none.
src/skillevaluator/tier3/loose_parts.py:1268
↓ 1 callersFunction_skill_metric_not_applicable
Scoreless ``skill_execution`` / ``skill_efficiency`` result. The default reason covers an arm without the skill. An eval case that gives the
src/skillevaluator/tier3/harbor/templates/eval.py:4110
↓ 1 callersFunction_skill_name
(results: list[ValidationResult], ae: dict[str, Any] | None)
src/skillevaluator/reporting/benchmark.py:1396
↓ 1 callersFunction_skill_owns_skill_md
Whether the skill's name, "the <name> skill's" or "the installed skill's" comes right before a SKILL.md. That SKILL.md is the skill's own; rename
src/skillevaluator/tier3/generate_dataset.py:524
↓ 1 callersFunction_skill_relative_path
Show a location relative to the skill folder instead of the machine path. *scan_root* is the folder the scanners ran on, when it is known. Withou
src/skillevaluator/reporting/html.py:1997
↓ 1 callersFunction_skill_root_variants
(candidate: str, skill_root: Path)
src/skillevaluator/validators/skillspector_scope.py:1156
↓ 1 callersFunction_skill_scoped_excluded_roots
Keep only generated-output roots located inside this skill tree.
src/skillevaluator/tier3/harbor/adapter.py:818
↓ 1 callersFunction_skill_script_names
File names of the skill's scripts in scripts/ and tools/, which the no-skill baseline does not have.
src/skillevaluator/tier3/evals_spec.py:895
↓ 1 callersFunction_skill_status
Do not paint a skill green when aggregate and child verdicts disagree.
ci/report_views.py:94
↓ 1 callersFunction_skill_workspace
(value: Any, config_path: Path)
src/skillevaluator/tier3/evals_config.py:832
↓ 1 callersMethod_skilleval_launch_env
(self)
src/skillevaluator/tier3/harbor/native_agents.py:137
↓ 1 callersMethod_skilleval_launch_match
Split Harbor's launch command, or return ``None`` for any other command.
src/skillevaluator/tier3/harbor/native_agents.py:144
↓ 1 callersMethod_skilleval_launch_shape
Extra checks on the launcher line beyond the launch pattern (Harbor's launch shape).
src/skillevaluator/tier3/harbor/native_agents.py:140
↓ 1 callersMethod_skilleval_rewrite_launcher
(self, launcher: str, match: re.Match[str])
src/skillevaluator/tier3/harbor/native_agents.py:134
↓ 1 callersMethod_skills
(self)
src/skillevaluator/tier3/loose_parts.py:868
↓ 1 callersFunction_skillspector_adapter_fix
The problem and actions for a failed SkillSpector LLM stage when ``SKILLSPECTOR_PROVIDER`` names its adapter. ``None`` when it does not: SkillSpe
src/skillevaluator/reporting/console_ui.py:1428
↓ 1 callersFunction_skillspector_baseline_for_scan
Point SkillSpector at the staged copy of a baseline that lives inside the skill. SkillSpector excludes only the exact file passed to ``--baseline
src/skillevaluator/validators/security.py:1048
↓ 1 callersFunction_skillspector_default_reasoning_effort
Return SkillEvaluator's reasoning effort when every configured SkillSpector model takes it. The effort is one global SkillSpector setting, so a s
src/skillevaluator/validators/security.py:752
↓ 1 callersFunction_skillspector_gaps_without_rows
Describe the gaps a report without its hook rows (:func:`_skillspector_without_hook_surface`) shows outside its rows. A gap is a limitation, a pa
src/skillevaluator/validators/security.py:2904
↓ 1 callersFunction_skillspector_gaps_without_rows_error
Describe a partial analysis whose only gaps outside the hook rows have no ledger row.
src/skillevaluator/validators/security.py:2952
↓ 1 callersFunction_skillspector_incomplete_files_by_reason
Count distinct files per ledger reason code, most common first.
src/skillevaluator/validators/security.py:1179
↓ 1 callersMethod_skillspector_issue_scopes
Return each issue's applicability, or ``None`` for an unrestricted scan.
src/skillevaluator/validators/security.py:5345
↓ 1 callersFunction_skillspector_key_kinds
*keys* split into plain name runs and other names, with their total and longest length. Each name run is given by its text before any trailing do
src/skillevaluator/validators/security.py:1966
↓ 1 callersFunction_skillspector_llm_stderr_failed
Detect SkillSpector LLM failures hidden behind exit 0 and clean JSON. SkillSpector 2.3.7 can catch provider/auth errors, log them to stderr, and
src/skillevaluator/validators/security.py:457
↓ 1 callersFunction_skillspector_markdown_labels
*text* with each bold or italic label outside a code span read as its text, and those labels as written. A label (:data:`_SKILLSPECTOR_MARKDOWN_L
src/skillevaluator/validators/security.py:2215
↓ 1 callersFunction_skillspector_missing_reference_warning
Describe accepted missing-reference ledger rows, naming the file and line of up to three of them.
src/skillevaluator/validators/security.py:2440
↓ 1 callersFunction_skillspector_operator_env
Return operator-set SkillSpector slot models, model registry and reasoning effort. ``SKILLSPECTOR_TEMPERATURE`` stays out: current reasoning mode
src/skillevaluator/validators/security.py:676
↓ 1 callersFunction_skillspector_partially_inspected_report
Return a partial report whose inventory, counters and analyzer accounting match *ledger_exceptions*. Files in *inspected* are fully inspected. Ev
tests/validators/test_security.py:347
↓ 1 callersFunction_skillspector_provenance
Return the validated SkillSpector LLM provenance records on ``results``.
src/skillevaluator/reporting/html.py:1842
↓ 1 callersFunction_skillspector_provider_env
Map public provider settings into an invocation-scoped SkillSpector environment. Also returns the SkillEvaluator provider the bridge resolved. It
src/skillevaluator/validators/security.py:688
↓ 1 callersFunction_skillspector_provider_retry_row
Return whether one ledger row is a provider-retry exhaustion on LLM analyzers only.
src/skillevaluator/validators/security.py:506
↓ 1 callersFunction_skillspector_report_llm_provenance
Return the validated ``metadata.llm_provenance`` of a SkillSpector JSON report, if any.
src/skillevaluator/validators/security.py:469
↓ 1 callersFunction_skillspector_runtime_limit_hint
Explain how to clear ``runtime_limit`` ledger exceptions, or ``None`` when there are none.
src/skillevaluator/validators/security.py:1190
↓ 1 callersFunction_skillspector_shell_expansions
*word* with its brace and parameter expansions read as ``*``, in time linear in its length. Everything from the first ``${`` to the last ``}`` af
src/skillevaluator/validators/security.py:2084
↓ 1 callersFunction_skillspector_shell_groups
*text* with the shell groups and ranges inside its words read as patterns, in time linear in its length. zsh reads ``(a)`` inside a word as ``a``
src/skillevaluator/validators/security.py:2028
↓ 1 callersFunction_skillspector_shell_word
*word* as glob patterns for what a POSIX shell, and cmd.exe or PowerShell, may expand it to. The work is linear in the length of *word*, and a gr
src/skillevaluator/validators/security.py:2118
↓ 1 callersFunction_skillspector_text_names
Whether *name* appears in *text* with no name character around it, after any trailing dots.
src/skillevaluator/validators/security.py:2425
↓ 1 callersFunction_skillspector_tilde_forms
Patterns for what the path part *part* may name when it has a ``~`` after its start. *part* itself is one. zsh ``EXTENDED_GLOB`` reads ``a~b`` as
src/skillevaluator/validators/security.py:2141
↓ 1 callersFunction_skillspector_versions
Return the distinct SkillSpector versions recorded on ``results``, in first-seen order.
src/skillevaluator/reporting/html.py:1831
↓ 1 callersFunction_skillspector_zsh_pattern
*pattern* with the zsh ``EXTENDED_GLOB`` operators ``^`` and ``#`` read, in time linear in its length. Outside a bracket expression, ``^`` negate
src/skillevaluator/validators/security.py:2168
↓ 1 callersFunction_skip
(message: str)
src/skillevaluator/cli.py:752
↓ 1 callersFunction_skip_arithmetic
Advance past an arithmetic expansion, where ``<<`` is a shift operator.
src/skillevaluator/tier3/eval_core/script_invocation.py:2111
↓ 1 callersFunction_skip_balanced_member
(tokens: list[tuple[str, str]], start: int, ends: dict[int, int])
src/skillevaluator/tier3/eval_core/codex_tool_call_normalizer.py:662
↓ 1 callersFunction_skip_js_regex
Return the end of one plausible regex literal or its invalid line.
src/skillevaluator/tier3/eval_core/codex_tool_call_normalizer.py:421
↓ 1 callersFunction_skip_name
The ``--skip`` value that leaves out the check producing *validator_name*.
src/skillevaluator/reporting/console_ui.py:1449
↓ 1 callersFunction_skip_transparent_prefixes
Advance past grouping tokens and wrappers that run the command after them.
src/skillevaluator/tier3/eval_core/script_invocation.py:2691
↓ 1 callersFunction_skip_views
(results: tuple[ValidationResult, ...])
src/skillevaluator/reporting/tier2_views.py:49
↓ 1 callersFunction_skipped
(reason: str, hint: str)
ci/request_gitlab_feedback.py:96
↓ 1 callersFunction_skipped_with_skill_activation
Activation summary for a baseline-only run, which never ran the skill.
src/skillevaluator/tier3/harbor/collector.py:314
↓ 1 callersFunction_slice_with_middle_marker
Compact *text* to fit *budget* by replacing the middle with *marker*.
src/skillevaluator/tier3/eval_core/llm_judge.py:936
↓ 1 callersFunction_slice_with_middle_marker
Compact *text* to fit *budget* by replacing the middle with *marker*.
src/skillevaluator/tier3/harbor/templates/eval.py:834
↓ 1 callersFunction_snapshot
(value: object, owner: str)
src/skillevaluator/reporting/plugin_evidence.py:51
↓ 1 callersFunction_snapshot
One bounded copy with inode-pinned, no-follow reads before scanners run. Scanners receive an immutable task-owned input view and cannot follow so
src/skillevaluator/tier1/plugin_checks.py:124
↓ 1 callersFunction_sorted_has_prefix
(sorted_values: list[str], prefix: str, *, exact: bool = False)
src/skillevaluator/validators/skillspector_scope.py:229
↓ 1 callersFunction_source_attempts
(reports: Path, current: int)
ci/run_tiers.py:669
↓ 1 callersFunction_source_distribution_root
(path: Path)
scripts/check_oss_boundary.py:941
↓ 1 callersFunction_source_files
One no-follow bounded listing for independent audits, with exclusions explicitly reported.
src/skillevaluator/plugin_inventory.py:249
↓ 1 callersFunction_source_folder_problem
Why a listed plugin's local source folder is unusable (no-follow), or ``None``. Returns ``(problem, cause)``. ``cause`` is ``None`` when the fold
src/skillevaluator/plugin_manifest.py:701
↓ 1 callersFunction_source_ranges_by_line
(node: ast.AST, lines: list[str])
scripts/check_oss_boundary.py:364
↓ 1 callersFunction_span_strictly_contains
(outer: _SourceSpan, inner: _SourceSpan)
scripts/check_oss_boundary.py:529
↓ 1 callersFunction_spawned_agent_id
The ``agentId`` of the subagent a parent ``Agent``/``Task`` call started, when its result says it. Claude Code returns the subagent's ``agentId``
src/skillevaluator/tier3/eval_core/plugin_signals.py:3229
↓ 1 callersFunction_special_scheme_slashes
Read ``\\`` as ``/`` before the query or fragment, as WHATWG does for special schemes.
src/skillevaluator/validators/url_policy.py:548
↓ 1 callersFunction_specificity
(selector: str)
tests/reporting/test_report_redesign.py:553
↓ 1 callersFunction_spelled_with_capitals
Whether a bare command word spells its executable with a capital letter, as prose ``HTTP`` or ``Https`` do. Prose has no path or ``.exe`` words,
src/skillevaluator/tier3/eval_core/checks.py:2183
↓ 1 callersFunction_split_chat_history
Split an OpenAI-style history into its leading system text and its conversation turns.
src/skillevaluator/inference/client.py:341
↓ 1 callersFunction_split_glued_syntax
Split compound syntax glued to the next word at ``idx``, in place. ``up(){`` becomes ``up()`` and ``{``, ``up()(curl`` becomes ``up()``, ``(``
src/skillevaluator/tier3/eval_core/checks.py:2521
↓ 1 callersFunction_split_harness_version
(value: str)
src/skillevaluator/cli.py:3242
↓ 1 callersFunction_split_recovered_failures
Separate active trial failures from timeouts a later attempt recovered.
src/skillevaluator/tier3/harbor/collector.py:6795
↓ 1 callersFunction_split_userinfo
``(user, password)``: the userinfo split at its first ``:`` outside a ``${...}`` reference.
src/skillevaluator/validators/url_policy.py:493
↓ 1 callersFunction_split_windows_container_file_source
Split a container file path while preserving drive-root semantics.
src/skillevaluator/tier3/harbor/secure_docker_environment.py:834
↓ 1 callersFunction_stage_bundle_files
Copy the bundle's single plugin files (bytes and mode kept), each contained in the plugin root. The copy never follows a link and never writes ov
src/skillevaluator/tier3/harbor/native_staging.py:225
↓ 1 callersFunction_stage_fallback_file
(manifest: _FileManifest, stage: Path)
src/skillevaluator/tier3/harbor/secure_copy.py:2291
↓ 1 callersMethod_stage_hooks
Merge every hook source into one census-wrapped ``hooks/hooks.json``, with one listing per source. Frontmatter hooks of a skill or command st
src/skillevaluator/tier3/plugin_native.py:1598
↓ 1 callersFunction_stage_manifests
( parent_descriptor: int, destination_name: str, manifests: tuple[_TreeManifest, ...], )
src/skillevaluator/tier3/harbor/secure_copy.py:1310
↓ 1 callersMethod_stage_mcp_servers
Write the staged ``.mcp.json``. It holds the runnable servers plus the ones that launch from the copied plugin files (Claude Code exp
src/skillevaluator/tier3/plugin_native.py:1574
↓ 1 callersFunction_stage_member_skills
The task skills projection of the member skills, with the emitter's own copy rules.
src/skillevaluator/tier3/loose_parts.py:1291
↓ 1 callersFunction_stage_native
(tmp_path: Path, compose: dict[str, Any])
tests/test_harbor_custom_compose_safety.py:46
↓ 1 callersFunction_stage_native_skills
Project exactly the layout consumed by the selected native adapter.
src/skillevaluator/tier3/plugin_execution.py:208
↓ 1 callersMethod_stage_plugin_configs
Write the applied ``settings.json`` and ``.lsp.json``, and list them and ``bin/``.
src/skillevaluator/tier3/plugin_native.py:1694
↓ 1 callersMethod_stage_skills
Copy the member skills Claude Code would not read where they are, and list every loaded skill.
src/skillevaluator/tier3/plugin_native.py:1553
↓ 1 callersFunction_stage_task_mcp
Add portable MCP declarations only to the private with arm.
src/skillevaluator/tier3/plugin_execution.py:244
↓ 1 callersMethod_stage_texts
List the subagents, commands, and output styles; an inline command becomes a staged file. The inline command's ``commands`` map entry in *man
src/skillevaluator/tier3/plugin_native.py:1659
↓ 1 callersFunction_staged_relative_link_path
(skill_name: str, raw_target: str)
src/skillevaluator/tier3/harbor/adapter.py:1140
↓ 1 callersFunction_staged_task_entry_id_map
Build an unambiguous mapping from staged task directory or task.name to [metadata].entry_id.
src/skillevaluator/tier3/harbor/collector.py:3550
↓ 1 callersMethod_stamp_manifest_metadata
(manifest_path: Path, root: Path, manifest_type: str, result: ValidationResult)
src/skillevaluator/validators/plugin_schema.py:125
↓ 1 callersFunction_stamp_tier1_applicability
Record on every result that the manifest checks do not apply to the target.
src/skillevaluator/tier1/commands.py:220
↓ 1 callersFunction_standard_aggregate_matches_harbor_strategy
Match FINAL, missing-as-zero MEAN, or either when the strategy is unknown.
src/skillevaluator/tier3/harbor/collector.py:3256
↓ 1 callersMethod_start_candidate
(self, prefix: str)
src/skillevaluator/tier3/harbor/stream_redaction.py:384
↓ 1 callersFunction_starter_evals_entry
(skill_path: Path, case_id: str)
src/skillevaluator/tier3/commands.py:105
↓ 1 callersFunction_stat_with_advanced_ctime
Copy every stat field and advance only ``st_ctime_ns``. ``os.stat_result(sequence, extras)`` fills the fields that are not part of the 10-ite
tests/test_results_location_legacy_compatibility.py:805
↓ 1 callersFunction_stat_with_advanced_ctime
Copy every stat field and advance only ``st_ctime_ns``. ``os.stat_result(sequence, extras)`` fills the fields that are not part of the 10-ite
tests/test_harbor_input_staging.py:32
← previousnext →7,801–7,900 of 24,741, ranked by callers