Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/NVIDIA/SkillEvaluator
/ functions
Functions
24,741 in github.com/NVIDIA/SkillEvaluator
⨍
Functions
24,741
◇
Types & classes
1,292
↳
Endpoints
703
↓ 1 callers
Function
_tier3_html_dimensions
Keep recorded dimensions and configured-but-unmeasured public dimensions visible.
src/skillevaluator/reporting/html.py:1056
↓ 1 callers
Function
_tier3_metric_comparison_view
Bound a custom-only Overview table to already assessed recorded metric values.
src/skillevaluator/reporting/html.py:1117
↓ 1 callers
Function
_tier3_metrics
(results: list[ValidationResult])
ci/summary_metrics.py:37
↓ 1 callers
Function
_tier3_results
(results: list[ValidationResult])
src/skillevaluator/reporting/benchmark.py:2142
↓ 1 callers
Function
_tier3_run_dir
(payload: dict)
src/skillevaluator/reporting/console_ui.py:1546
↓ 1 callers
Function
_tier3_score_reasons
Keep per-metric reasons keyed like ``_score_mapping`` so each matches its score cell.
src/skillevaluator/reporting/html.py:505
↓ 1 callers
Function
_tier3_source_validation_rows
Count the failed Tier 3 source checks and, unless the CLI report lists them, show them.
src/skillevaluator/reporting/console_ui.py:1525
↓ 1 callers
Function
_tier3_string_list
Return a small list of safe strings from a canonical JSON array.
src/skillevaluator/reporting/console_ui.py:1049
↓ 1 callers
Function
_tier_status
( tier: str, results: list[ValidationResult], ae: dict[str, Any] | None, benchmark_policy: dic
src/skillevaluator/reporting/benchmark.py:1959
↓ 1 callers
Function
_tier_status_rows
Return Tier rows from rendered root-level Markdown tables.
scripts/ci/check_public_benchmarks.py:1578
↓ 1 callers
Function
_time
(value)
ci/feedback.py:101
↓ 1 callers
Function
_timestamp
(value: str)
ci/publish_reports.py:689
↓ 1 callers
Function
_timestamp
(value)
ci/verify_provenance.py:99
↓ 1 callers
Function
_token_chart
(raw: object)
src/skillevaluator/reporting/telemetry_view.py:256
↓ 1 callers
Function
_token_counter
Return one browser-safe token count, preserving unavailable as null.
src/skillevaluator/evaluation/tier3_report.py:150
↓ 1 callers
Function
_token_usage_fits
(payload: dict[str, Any])
src/skillevaluator/evaluation/tier3_report.py:1821
↓ 1 callers
Function
_token_usage_lines
Show measured token usage per agent and per case; it is not scored.
src/skillevaluator/reporting/benchmark.py:1547
↓ 1 callers
Function
_tokenize
(clause: str)
src/skillevaluator/evaluation/dataset_case_removal.py:216
↓ 1 callers
Function
_toml_string
A TOML basic string. JSON escapes are not TOML: an emoji would become a surrogate pair.
src/skillevaluator/tier3/plugin_native.py:1286
↓ 1 callers
Function
_tool_call_productivity
Classify one tool call as ``productive``, ``wasted`` or ``neutral`` (not scored).
src/skillevaluator/tier3/eval_core/checks.py:8321
↓ 1 callers
Function
_tool_call_refs
(traj: dict[str, Any])
src/skillevaluator/tier3/eval_core/atif_helpers.py:707
↓ 1 callers
Function
_tool_call_seconds
Measured wall time for each tool call on the step, or ``None`` if any call is unmeasured.
src/skillevaluator/tier3/harbor/trajectory_timing.py:241
↓ 1 callers
Function
_tool_efficiency_text
Shell text for a tool call target. Codex records some shell calls as an argv list such as ``["bash", "-lc", "find / -name x"]``. Joining it t
src/skillevaluator/tier3/eval_core/checks.py:8276
↓ 1 callers
Function
_tool_entries
Claude ``tools``-style values: a YAML list or a comma list (commas inside ``(...)`` kept).
src/skillevaluator/tier3/plugin_native.py:1880
↓ 1 callers
Function
_tool_name
(tool_call: Mapping[str, Any])
src/skillevaluator/tier3/eval_core/plugin_signals.py:1740
↓ 1 callers
Function
_tool_observation_refs
(traj: dict[str, Any])
src/skillevaluator/tier3/eval_core/atif_helpers.py:719
↓ 1 callers
Function
_tool_searches_root
Recognize explicit root search operands, not '/' inside a pattern. This is bounded top-level command classification. It does not execute or i
src/skillevaluator/tier3/eval_core/checks.py:7975
↓ 1 callers
Function
_tool_step
(index: int, name: str, arguments: dict)
tests/tier3/test_tool_productivity.py:547
↓ 1 callers
Function
_tools_reference_is_called
(tokens: list[tuple[str, str]], end: int, bracket_ends: dict[int, int])
src/skillevaluator/tier3/eval_core/codex_tool_call_normalizer.py:667
↓ 1 callers
Function
_top_findings
(findings: list[Finding], *, limit: int)
src/skillevaluator/reporting/benchmark.py:2157
↓ 1 callers
Function
_touches_tier3_platform
(path: bytes)
scripts/classify_ci_changes.py:69
↓ 1 callers
Function
_tracked_paths
(root: Path)
scripts/check_oss_boundary.py:779
↓ 1 callers
Function
_trajectory
(variant: str)
tests/tier3/test_private_runtime_report_acceptance.py:64
↓ 1 callers
Function
_trajectory_agent_runtime_failure_reason
(trajectory: Any)
src/skillevaluator/tier3/harbor/collector.py:754
↓ 1 callers
Function
_translate_backend
(completion: Any, translator: Any, *translator_args: Any)
src/skillevaluator/tier3/harbor/nvidia_build_bridge.py:430
↓ 1 callers
Function
_tree
(api, root, commit)
ci/benchmark_publish.py:54
↓ 1 callers
Function
_tree_checks
()
src/skillevaluator/tier1/plugin_checks.py:48
↓ 1 callers
Function
_tree_contains_artifact_dirs
Return True when the SkillSpector scan tree needs staging.
src/skillevaluator/validators/security.py:783
↓ 1 callers
Function
_tree_digest
Hash relative file content and file/directory rwx modes without following links.
src/skillevaluator/tier3/loose_parts.py:1326
↓ 1 callers
Function
_tree_size
(directory: Path)
ci/publish_reports.py:821
↓ 1 callers
Function
_trial
(job: Path, name: str, result: dict | None)
tests/reporting/test_tier3_validate_cases.py:194
↓ 1 callers
Function
_trial_census
( native: Any, agent: str, logs: Path, declared: Any, plan: Mapping[str, Any], *,
src/skillevaluator/tier3/plugin_execution.py:561
↓ 1 callers
Function
_trial_codex_private_dirs_ignore
Return a secure-copy ignore callback that drops local Codex runtime dirs. Harbor's Codex agent links ``codex-home/auth.json`` to the credential i
src/skillevaluator/tier3/harbor/runner.py:4416
↓ 1 callers
Function
_trial_derived_job_stats
Return the job counters Harbor derives from trials, independent of order. Harbor fills job-level ``metrics`` and ``pass_at_k`` after aggregating
src/skillevaluator/tier3/harbor/collector.py:6764
↓ 1 callers
Function
_trial_error_summary
(trial_dir: Path)
src/skillevaluator/tier3/harbor/collector.py:1434
↓ 1 callers
Function
_trial_failures
(agent: str, data: Mapping[str, Any])
src/skillevaluator/tier3/result_display.py:334
↓ 1 callers
Function
_trial_keys
(trials: list[dict[str, Any]])
src/skillevaluator/reporting/dashboard_data.py:186
↓ 1 callers
Function
_trial_list
(raw: object)
src/skillevaluator/reporting/trial_charts.py:35
↓ 1 callers
Function
_trial_negative_control_failure
Interpret one trial, preferring strict invocation evidence over legacy scores.
src/skillevaluator/evaluation/insights_judge.py:396
↓ 1 callers
Function
_trial_output_tokens
Return proof that the agent produced output before a trial ended, or None. Harbor's ``agent_result.n_output_tokens`` comes first; otherwise any
src/skillevaluator/tier3/harbor/collector.py:7318
↓ 1 callers
Function
_trial_resumes_step_trajectories
Read Harbor 0.22's serialized agent resume flag without coercion.
src/skillevaluator/tier3/harbor/collector.py:2808
↓ 1 callers
Function
_trial_score_reasons
Return one bounded grader reason per failing metric of a logical trial. A logical trial can span several reward rows. The reason comes from the
src/skillevaluator/evaluation/tier3_report.py:3764
↓ 1 callers
Function
_trusted_pins
(variable: str, *, optional=False)
ci/publish_reports.py:513
↓ 1 callers
Function
_trusted_trial_skill_invoked
Derive trusted invocation from the topology-authorized materialized ATIF.
src/skillevaluator/tier3/harbor/collector.py:6393
↓ 1 callers
Function
_trusted_wrappers
()
ci/publish_reports.py:606
↓ 1 callers
Function
_type_ok
(value: Any, expected: str)
src/skillevaluator/tier3/eval_core/plugin_signals.py:3845
↓ 1 callers
Function
_typed_metadata
Validate nested containers dereferenced by native report templates.
src/skillevaluator/reporting/retained.py:73
↓ 1 callers
Function
_unescaped
Return *text* with one level of string escapes decoded, and where each of its characters starts in *text*. The starts end with len(text), so the
src/skillevaluator/tier3/eval_core/checks.py:4045
↓ 1 callers
Function
_union_ms
Wall-clock length of the union of ``[start, end]`` intervals.
src/skillevaluator/tier3/harbor/trajectory_timing.py:150
↓ 1 callers
Function
_unique
(items: list[dict[str, Any]])
src/skillevaluator/reporting/insights_view.py:65
↓ 1 callers
Function
_unique_id
(base: str, seen: set[str])
src/skillevaluator/tier3/plugin_prepare.py:833
↓ 1 callers
Function
_unique_mapping
(pairs: list[tuple[Any, Any]])
src/skillevaluator/tier3/dataset_migration.py:34
↓ 1 callers
Function
_unique_name
(base: str, used: set[str])
src/skillevaluator/reporting/junit_reporter.py:184
↓ 1 callers
Function
_unit_is_complete
Whether ``unit`` ends a complete command, so the newline after it ends what the shell parses before running it. Every compound it opens is cl
src/skillevaluator/tier3/eval_core/script_invocation.py:1686
↓ 1 callers
Function
_unit_score
(value: object)
src/skillevaluator/tier3/harbor/templates/metric.py:74
↓ 1 callers
Function
_unpack
(data: Any)
src/skillevaluator/tier3/dataset_migration.py:71
↓ 1 callers
Function
_unparsed_url
A malformed URL for reports: the query, fragment, and userinfo (through the authority's last ``@``) removed.
src/skillevaluator/validators/url_policy.py:873
↓ 1 callers
Function
_unprotect_stored_key
Decode DPAPI-protected Windows bytes or return POSIX bytes unchanged.
src/skillevaluator/tier3/output_provenance.py:289
↓ 1 callers
Function
_unquoted_shell_text
(command: str)
src/skillevaluator/tier3/harbor/local_environment.py:301
↓ 1 callers
Function
_unrelated_negation_advice
A noun between the negator (and its modifiers) and the verb: the negator governs the noun, not the removal.
tests/test_tier3_case_summary.py:428
↓ 1 callers
Function
_unrelated_negation_advice
A noun between the negator (and its modifiers) and the verb: the negator governs the noun, not the removal.
tests/evaluation/test_judges.py:926
↓ 1 callers
Function
_unrepresented_insight_steps
Keep next steps unless a failed-run card already displays all their evidence.
src/skillevaluator/reporting/html.py:1625
↓ 1 callers
Function
_unresolved_artifact_path
Return the absolute path a finding names, without following its symlinks. ``None`` when the path has a ``..`` segment: after a symlinked director
src/skillevaluator/reporting/sarif_reporter.py:138
↓ 1 callers
Function
_unresolved_reference_fixes
Explain a SkillSpector scan that ran but was not accepted because a file reference didn't resolve.
src/skillevaluator/reporting/console_ui.py:1288
↓ 1 callers
Function
_unsafe_command
(severity: Severity, line_number: int)
tests/reporting/test_reporters.py:993
↓ 1 callers
Function
_unsafe_trial_directory_reason
Return a bounded reason when an expected Harbor trial root is unsafe.
src/skillevaluator/tier3/harbor/collector.py:1137
↓ 1 callers
Method
_unsafe_write_target
(self, command: str, *, exec_env: dict[str, str])
src/skillevaluator/tier3/harbor/local_environment.py:1699
↓ 1 callers
Function
_unscanned_server_files
HIGH when an MCP server runs a plugin file from a folder that Tier 1 whole-tree scans skip. The server's launch names plugin files (:func:`_serve
src/skillevaluator/plugin_mcp.py:551
↓ 1 callers
Function
_unscanned_skill_finding
HIGH for a skill a client loads from a declared skills folder that Tier 1 never scans. The skill sits deep in an ``evals/``, ``results/``, or ``v
src/skillevaluator/plugin_member_paths.py:200
↓ 1 callers
Function
_unscored_attempt_usage
Read the token counters of an attempt that has no usable reward. The attempt is not scored, but it ran and used tokens. Its case is the one t
src/skillevaluator/tier3/harbor/report_data.py:790
↓ 1 callers
Function
_unscored_attempts_before_first_pass
Count attempt slots up to the first pass that produced no scored result. Stop-on-pass skips only the attempts after the first pass. A slot at or
src/skillevaluator/tier3/harbor/collector.py:5108
↓ 1 callers
Function
_unscored_trial_case_id
Return the case of a Harbor trial without a scored reward, from the task it ran. The task comes from the trial's ``result.json``, or from its ``c
src/skillevaluator/tier3/harbor/collector.py:1523
↓ 1 callers
Function
_unsettle_every_binding
Text this walk cannot read may bind or unbind any name: none stays settled.
src/skillevaluator/tier3/eval_core/script_invocation.py:1355
↓ 1 callers
Function
_unsettled_cd_operand
Return whether a shell ``cd`` *operand* keeps quoting the lexical reading does not remove. One surrounding pair of quotes is read as the shell re
src/skillevaluator/tier3/eval_core/checks.py:1351
↓ 1 callers
Method
_unshipped_code
Findings for code a hook runs that the plugin does not ship (and that is not remote code). A package or image with a floating version or tag
src/skillevaluator/plugin_component_risk.py:3665
↓ 1 callers
Method
_unsupported_rows
(self, source: NativePluginSource)
src/skillevaluator/tier3/plugin_native.py:1513
↓ 1 callers
Function
_unsupported_security_result
(result: Mapping[str, Any])
src/skillevaluator/tier3/attempt_verdict.py:80
↓ 1 callers
Method
_untaken
(self, index: int)
src/skillevaluator/tier3/eval_core/checks.py:4544
↓ 1 callers
Function
_untrusted_structural_heading_lines
Return root heading lines whose rendered structure is ambiguous.
scripts/ci/check_public_benchmarks.py:1321
↓ 1 callers
Function
_untrusted_tls_context
(tmp_path: Path)
tests/test_retry_provider_transport.py:197
↓ 1 callers
Function
_unverified_message
(manifest: str)
tests/validators/test_skillspector_hook_surface.py:174
↓ 1 callers
Function
_unwrap_kubeconfig_node
Recursively unwrap Kubernetes ConfigNode wrappers while absolutizing relative file paths.
src/skillevaluator/tier3/harbor/kubeconfig.py:49
↓ 1 callers
Function
_unwrapped_texts
The tool texts inside an MCP content wrapper that opens ``text``, else ``[]``. Codex prints an MCP result as ``[{"type":"text","text":"..."}]`` a
src/skillevaluator/tier3/eval_core/plugin_signals.py:1804
↓ 1 callers
Method
_update
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateAnimationTarget
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateDataset
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateDatasets
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateHiddenIndices
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateLayout
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callers
Method
_updateMetasets
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
← previous
next →
8,001–8,100 of 24,741, ranked by callers