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
↓ 2 callers
Function
_escaped_twice
(text: str)
tests/tier3/test_security_finding_evidence.py:1232
↓ 2 callers
Function
_estimate_skill
(tmp_path)
tests/validators/test_quality_score.py:2104
↓ 2 callers
Function
_evaluation_mode
Classify the arms a run executed: ``both``, ``with_skill_only`` or ``baseline_only``. Read from the collector's deliberate per-arm ``execution_st
src/skillevaluator/evaluation/tier3_report.py:4748
↓ 2 callers
Method
_eventHandler
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_event_detail
(self, event: ProgressEvent)
src/skillevaluator/tier3/harbor/progress.py:397
↓ 2 callers
Function
_evidence
(run_dir: Path, trial_dir: Path | None, *, run_error: bool)
src/skillevaluator/tier3/case_summary.py:1223
↓ 2 callers
Function
_evidence
(payload: dict, record: dict)
ci/tier_manifest.py:320
↓ 2 callers
Method
_evidence
(self, rel: PurePosixPath, unanalyzed: list[str])
src/skillevaluator/plugin_component_risk.py:3282
↓ 2 callers
Function
_evidence_anchors
Give every check and finding card in the skill view a stable, unique id. Returns ``(findings, checks)``: ``findings[skill][check]`` lists one id
src/skillevaluator/reporting/html.py:2020
↓ 2 callers
Function
_evidence_case_ids
(item: dict[str, Any])
src/skillevaluator/evaluation/tier3_report.py:4197
↓ 2 callers
Function
_exact_publication_text
Preserve a filesystem identity only when its exact NFC text is safe.
src/skillevaluator/reporting/benchmark.py:2297
↓ 2 callers
Function
_exception_cause
Follow an explicit cause, otherwise the unsuppressed implicit context.
src/skillevaluator/inference/retry.py:119
↓ 2 callers
Function
_excluded
(name: str, relative: Path)
ci/diagnostics.py:111
↓ 2 callers
Method
_exec_pod_stream_command
Execute a command via Kubernetes WebSocket exec stream, draining output and returning returncode.
src/skillevaluator/tier3/harbor/gke_environment.py:1051
↓ 2 callers
Function
_exec_result
A shell result inside the status header Harbor puts above Codex ``exec`` output.
tests/tier3/test_plugin_signals.py:67
↓ 2 callers
Function
_exec_step
(timestamp: str, call_id: str)
tests/tier3/test_trajectory_timing.py:399
↓ 2 callers
Function
_execute_vertex_http_probe
Execute a Vertex HTTP probe request and map status/exceptions to ModelProbeResult.
src/skillevaluator/tier3/harbor/runtime_preflight.py:1605
↓ 2 callers
Function
_execution_error_details
Normalize bounded execution-error detail metadata around visible errors.
src/skillevaluator/tier3/harbor/report_data.py:606
↓ 2 callers
Method
_existing_file
(path: str | None)
src/skillevaluator/tier3/harbor/runner.py:5438
↓ 2 callers
Function
_existing_windows_target_usage
Conservatively estimate transactional copies of an existing target.
src/skillevaluator/tier3/harbor/secure_docker_environment.py:1137
↓ 2 callers
Function
_exit_one_observations
The same failed command as a plain status line, a Harbor Claude Code result and a Codex result.
tests/tier3/test_error_recovery.py:1194
↓ 2 callers
Function
_expand
(path: str | Path)
src/skillevaluator/tier3/results_location.py:50
↓ 2 callers
Function
_expand_url_defaults
An MCP URL as Claude Code expands it when no variable is set. ``${NAME:-default}`` becomes its default. ``${NAME}``, ``${NAME:-}``, and Curso
src/skillevaluator/validators/mcp_static.py:878
↓ 2 callers
Function
_expected_contract
(grading_mode: str | None)
tests/test_harbor_unscored_grading_policy.py:26
↓ 2 callers
Function
_expected_typed_runtime_reason
(exception_type: str, message: str)
tests/test_harbor_collector_runtime_failures.py:49
↓ 2 callers
Function
_expose_tree_root_exact
Validate while root-private, then expose only by changing root mode.
src/skillevaluator/tier3/harbor/secure_copy.py:1274
↓ 2 callers
Function
_extend_task_input_projection
Add the final evaluator-owned input replacement to generated Dockerfiles.
src/skillevaluator/tier3/harbor/adapter.py:3920
↓ 2 callers
Function
_extract_agent_final_message
Extract the agent's last substantive message from the trajectory.
src/skillevaluator/tier3/generate_dataset.py:1481
↓ 2 callers
Method
_extract_contributors
Extract contributor summary by mapping authors to their content items. Searches for author information in: 1. Success details metadat
src/skillevaluator/reporting/html.py:2773
↓ 2 callers
Function
_extract_custom_metrics
(reward: dict[str, Any])
src/skillevaluator/tier3/harbor/templates/custom_grader_runner.py:233
↓ 2 callers
Function
_extract_json
Extract a JSON payload from LLM response text. Tolerates markdown fences and prose around exactly one valid bounded JSON container. Multiple
src/skillevaluator/tier3/eval_core/llm_judge.py:311
↓ 2 callers
Method
_extract_spdx_from_file
Extract SPDX license identifier from file header.
src/skillevaluator/validators/license.py:667
↓ 2 callers
Function
_fail_closed_invalid_reward_numbers
Replace invalid numbers and attach a bounded diagnostic to the reward.
src/skillevaluator/tier3/harbor/collector.py:3946
↓ 2 callers
Function
_failed
True when a grader sub-result explicitly records a failure.
src/skillevaluator/evaluation/tier3_findings.py:142
↓ 2 callers
Function
_failed_judge_sidecar_paths
Return bounded sidecar paths plus a safe reason when the scan is incomplete. The candidate bound applies to sidecars that actually exist. Empty v
src/skillevaluator/tier3/harbor/collector.py:984
↓ 2 callers
Function
_failed_operands
Operands that a shell error line in ``observation`` names as unreadable.
src/skillevaluator/tier3/eval_core/plugin_signals.py:2682
↓ 2 callers
Function
_failed_run
Return a run that failed through a failed peer or a backend error after scoring. With ``budget``, the run's budget also left planned attempts not
tests/reporting/test_tier3_completed_agent_docs.py:464
↓ 2 callers
Function
_failure_list_metadata
Describe exact failure cardinality beside a bounded published sample.
src/skillevaluator/tier3/harbor/collector.py:477
↓ 2 callers
Function
_failure_text
(row: dict[str, Any])
src/skillevaluator/tier3/nemo_gym/results.py:114
↓ 2 callers
Function
_fake_git_output
Return a check_output stand-in that emits raw git bytes decoded like a cp1252 locale.
tests/utils/test_helpers.py:208
↓ 2 callers
Function
_fake_streamlit_spec
Fake only Streamlit's lookup; coverage and pytest call find_spec for other modules.
tests/reporting/test_dashboard_cli.py:46
↓ 2 callers
Method
_feed
(self, text: str, *, final: bool)
src/skillevaluator/tier3/harbor/stream_redaction.py:517
↓ 2 callers
Function
_field_string_leaves
Read at most 1000 values without copying huge containers or expanding aliases.
src/skillevaluator/tier3/evals_spec.py:918
↓ 2 callers
Function
_file_uri
Return an absolute ``file:`` URI for a path outside the workspace.
src/skillevaluator/reporting/sarif_reporter.py:127
↓ 2 callers
Function
_file_write_reward
(entry_id="case-a")
tests/tier3/test_suggestion_grounding.py:180
↓ 2 callers
Function
_fill
Substitute ``{fact}`` placeholders; numbers get thousands separators.
src/skillevaluator/harness/overlay.py:985
↓ 2 callers
Function
_final_message
(trial_dir: Path, run_dir: Path)
src/skillevaluator/tier3/case_summary.py:999
↓ 2 callers
Function
_finalize_file_exact
( parent: int, node: _StagedNode, manifest: _FileManifest, )
src/skillevaluator/tier3/harbor/secure_copy.py:2435
↓ 2 callers
Function
_finalize_summary
Finalize counts and stable ordering for a set of finding groups. ``reported_occurrences`` counts every location the raw records report, so ``
src/skillevaluator/reporting/finding_summary.py:84
↓ 2 callers
Method
_find_exact_child
Return a child path whose on-disk basename exactly matches ``filename``.
src/skillevaluator/validators/base.py:177
↓ 2 callers
Function
_find_skills
Skill folders under ``folder``, symlinked ones included (Claude Code and Codex load those).
src/skillevaluator/harness/overlay.py:635
↓ 2 callers
Function
_finding
(target: EndpointTarget, check: str, message: str, severity: Severity = Severity.MEDIUM)
src/skillevaluator/validators/endpoint_resolution.py:165
↓ 2 callers
Method
_finding_legacy_lines
Return the legacy error and warning lines that only repeat a structured finding.
src/skillevaluator/models/result.py:638
↓ 2 callers
Function
_finding_metric_names
(rewards: list[dict[str, Any]])
src/skillevaluator/tier3/harbor/report.py:251
↓ 2 callers
Function
_finding_severity_value
(finding: Finding)
src/skillevaluator/reporting/sarif_reporter.py:55
↓ 2 callers
Function
_findings_artifact_path
Return a findings path only when its parent remains inside results_dir.
src/skillevaluator/tier3/harbor/report.py:87
↓ 2 callers
Function
_fingerprint
(text: str)
src/skillevaluator/embedding/registry.py:718
↓ 2 callers
Function
_fingerprint
(profile, root_relative: str)
tests/deduplication/plugin/test_openclaw_profile.py:109
↓ 2 callers
Function
_finite_report_number
(value: object)
src/skillevaluator/reporting/markdown.py:80
↓ 2 callers
Function
_first_positional
(tokens: list[str], value_flags: frozenset[str])
src/skillevaluator/validators/mcp_static.py:2031
↓ 2 callers
Function
_first_string
(args: Mapping[str, Any], keys: Sequence[str])
src/skillevaluator/tier3/eval_core/plugin_signals.py:2250
↓ 2 callers
Function
_fit_row
Lay a row out in ``room`` columns so nothing wraps under the label column.
src/skillevaluator/reporting/console_ui.py:469
↓ 2 callers
Function
_fix_first
Pick the few places to fix first, most severe first, one entry per location. A check that could not finish comes first, with the scanner's own re
src/skillevaluator/reporting/html.py:2052
↓ 2 callers
Function
_fixed_size_chunks
Fall back to fixed-size overlapping windows.
src/skillevaluator/embedding/client.py:391
↓ 2 callers
Function
_fixture_secret
Build committed fake secrets from pieces so static scanners do not flag them.
tests/tier3/test_github_token_redaction.py:37
↓ 2 callers
Function
_fixture_secret
Build committed fake secrets from pieces so static scanners do not flag them.
tests/tier3/test_checks_secret_patterns.py:23
↓ 2 callers
Function
_flat
Lowercase text with Markdown code marks dropped and whitespace collapsed.
tests/tier3/test_dataset_baseline_fairness.py:1037
↓ 2 callers
Method
_flush_static
Print blocks once, in order, as they complete.
src/skillevaluator/reporting/console_ui.py:867
↓ 2 callers
Function
_follows_path_separator
Whether the character before ``index`` ends an earlier path or option. The colon of a drive letter (``C:\\``) does not: it belongs to the path.
src/skillevaluator/evaluation/tier3_feedback_chat.py:497
↓ 2 callers
Function
_foreign_namespace
Whether ``<namespace>:<name>`` names another plugin's component (unknown plugin name: never).
src/skillevaluator/tier3/eval_core/plugin_signals.py:2732
↓ 2 callers
Function
_format_dataset_size_summary
( status: str, distinct_cases: int, negative_cases: int | None, reason: str | None = None, )
src/skillevaluator/tier3/dataset_sufficiency.py:15
↓ 2 callers
Function
_format_percentage
(value: int, decimal_places: int)
src/skillevaluator/reporting/benchmark.py:1947
↓ 2 callers
Function
_format_units
(value: int, decimal_places: int)
src/skillevaluator/reporting/benchmark.py:1941
↓ 2 callers
Function
_freeze_gate_clock
( monkeypatch: pytest.MonkeyPatch, instant: datetime, )
tests/test_public_benchmark_gate.py:39
↓ 2 callers
Function
_frontmatter_root_kind
Classify the root node of the text between the ``---`` lines. Returns ``"mapping"`` for a plain YAML mapping, ``"invalid"`` when the YAML is
src/skillevaluator/deduplication/utils/skill_collector.py:351
↓ 2 callers
Method
_gate
(self)
src/skillevaluator/reporting/console_ui.py:722
↓ 2 callers
Function
_gate_label
(payload: dict)
ci/tier_manifest.py:352
↓ 2 callers
Function
_gate_policy
(payload: dict)
ci/tier_manifest.py:360
↓ 2 callers
Function
_gated
(result, tier, blocking=True)
tests/reporting/test_report_navigation.py:127
↓ 2 callers
Function
_generate_harbor_tasks_into
Generate Harbor task directories inside a private output directory. Args: skill_path: Path to the skill directory (contains SKILL.md, scr
src/skillevaluator/tier3/harbor/adapter.py:6199
↓ 2 callers
Function
_generate_suggestions_structured
Generate max 4 LLM-powered structured suggestion objects for skill developers. Each object has keys: ``suggestion`` (str), ``dimension`` (metric
src/skillevaluator/tier3/harbor/report.py:819
↓ 2 callers
Function
_generate_with_llm
Generate test cases using LLM for more natural questions.
src/skillevaluator/tier3/generate_dataset.py:684
↓ 2 callers
Function
_generated_texts
Commands with and without credentials, escaped up to three times, some with stray escapes and quotes.
tests/tier3/test_security_finding_evidence.py:1626
↓ 2 callers
Function
_generation_record
( skill_path: Path, *, mode: str, grounding: str, skill: dict[str, Any], prompt_path:
src/skillevaluator/tier3/generate_dataset.py:1787
↓ 2 callers
Method
_geojson_coordinate_lines
Return 1-based numbers of lines inside a GeoJSON geometry's ``"coordinates"`` array. A pretty-printed ``Polygon`` or ``Multi*`` geometry open
src/skillevaluator/validators/security.py:6293
↓ 2 callers
Method
_getAxis
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getCircumference
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getLabelCapacity
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getRingWeight
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getRingWeightOffset
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getRotation
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getSharedOptions
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getStackCount
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Method
_getStacks
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 2 callers
Function
_get_metadata_probe_image
Return the trusted container image used for the GKE metadata isolation probe.
src/skillevaluator/tier3/harbor/gke_environment.py:177
↓ 2 callers
Function
_git
(repository: Path, *arguments: str)
ci/smoke_image.py:152
↓ 2 callers
Function
_git_ref
The branch or tag a git/URL spec names, or ``None`` when it names none. npm reads the ref after ``#`` (``github:o/r#main``, ``git+https://h/o/r.g
src/skillevaluator/validators/mcp_static.py:2347
↓ 2 callers
Function
_github_run
(run, repository, run_id, attempt, *, events=frozenset({"pull_request"}))
ci/feedback.py:308
↓ 2 callers
Function
_gitlab_inherited
Resolve key as extends does: the job's own value, even null, wins, then the last parent defining it.
src/skillevaluator/_ci_onboarding_all_tiers.py:922
↓ 2 callers
Function
_gitleaks_allowlist_paths
()
tests/validators/test_artifact_dir_exclusion.py:358
← previous
next →
3,901–4,000 of 24,741, ranked by callers