MCPcopy Create free account

hub / github.com/NVIDIA/SkillEvaluator / functions

Functions24,741 in github.com/NVIDIA/SkillEvaluator

↓ 1 callersFunctionskillevaluator_manifest_fields
The manifest fields SkillEvaluator fails, for the field-level parity comparison. ``blocking`` lists every top-level manifest field with a CRITICA
src/skillevaluator/validators/claude_plugin_validate.py:684
↓ 1 callersFunctionskillevaluator_manifest_verdict
SkillEvaluator's own verdict from the plugin schema result: ``passed``, ``failed``, or ``None``.
src/skillevaluator/validators/claude_plugin_validate.py:649
↓ 1 callersMethodskills
(self)
src/skillevaluator/plugin_discovery.py:618
↓ 1 callersFunctionskills_changed
Return False only when a pull or merge request changes no file under a skill root. The paid jobs then skip and the keyless Tier 1 scan is the gat
ci/select_changed_skills.py:329
↓ 1 callersFunctionskillspector_report_is_truncated
Return whether a report serializes fewer occurrence records than findings it counted. SkillSpector bounds its output by occurrence record, so fin
src/skillevaluator/validators/skillspector_scope.py:435
↓ 1 callersFunctionskillspector_report_serializes_every_finding
Return whether every finding a report counted has exactly one serialized record. Each issue occurrence (or an issue without occurrences) is one r
src/skillevaluator/validators/skillspector_scope.py:450
↓ 1 callersFunctionskip
(reason: str, sid: Any = sid, record: int = record, **extra: Any)
src/skillevaluator/tier3/trace_import.py:260
↓ 1 callersFunctionslug_from_remote_url
Extract the ``<group>/<repo>`` slug from a git-remote URL. :func:`local_repo_slug` passes the raw ``origin`` URL of a hosted remote (see :dat
src/skillevaluator/plugin_dependencies.py:131
↓ 1 callersFunctionsmoke
(image: str, output: Path)
ci/smoke_image.py:377
↓ 1 callersFunctionsmoke
(**env: str)
tests/test_harbor_runtime_skill_isolation.py:4395
↓ 1 callersFunctionsnapshot
()
tests/test_version_comparison.py:199
↓ 1 callersFunctionsnapshot
()
tests/test_tier3_history.py:191
↓ 1 callersFunctionsources
(trials: list[dict[str, Any]])
src/skillevaluator/tier3/harbor/resource_usage.py:148
↓ 1 callersFunctionspecifier_pin
What one version specifier pins: ``==V`` and uv's ``@V`` when ``V`` is a PEP 440 version, and ``===V``. The MCP pinning check and the dependency
src/skillevaluator/validators/mcp_static.py:2289
↓ 1 callersMethodspend
Take *cost* from the budget; ``False`` once it is overdrawn.
src/skillevaluator/validators/security.py:1959
↓ 1 callersFunctionss
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctionstage
(skill, output, **kwargs)
tests/tier3/test_decoy_workspace.py:313
↓ 1 callersFunctionstage_frontmatter_hooks
Rewrite each skill or command file whose frontmatter declares hooks, with census-wrapped handlers. Claude Code runs frontmatter hooks only while
src/skillevaluator/tier3/plugin_native.py:726
↓ 1 callersFunctionstage_hint_row
A muted single-line hint shown while a tier is running.
src/skillevaluator/reporting/console_ui.py:902
↓ 1 callersFunctionstaged_paths
(case_id: str)
tests/test_harbor_input_staging.py:287
↓ 1 callersMethodstart
Track the run plan; once it names the run directory, poll case results.
src/skillevaluator/reporting/console_ui.py:162
↓ 1 callersMethodstart
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:7
↓ 1 callersMethodstart
(self, plan: Tier3RunPlan)
src/skillevaluator/tier3/harbor/progress.py:912
↓ 1 callersMethodstart_progress
Start the shared lifecycle before a plan (for the validation tiers).
src/skillevaluator/tier3/harbor/progress.py:222
↓ 1 callersMethodstart_progress
(self)
src/skillevaluator/tier3/harbor/progress.py:463
↓ 1 callersMethodstart_task
Mark a task as in_progress.
src/skillevaluator/tier3/reference_skills/task-list/scripts/tasks.py:198
↓ 1 callersFunctionstarter_guidance
(skill_path: Path)
src/skillevaluator/tier3/dataset_coverage.py:89
↓ 1 callersMethodstat
(self, *, follow_symlinks: bool = True)
tests/test_harbor_runtime_skill_isolation.py:291
↓ 1 callersFunctionstate_line
(identity: str, number: int)
tests/test_harbor_secure_docker_environment.py:2307
↓ 1 callersFunctionstep_agent_dirs
(trial_name: str)
src/skillevaluator/tier3/harbor/runner.py:4453
↓ 1 callersFunctionstep_number_from_url
(url: str)
src/skillevaluator/reporting/harbor_viewer.py:271
↓ 1 callersFunctionstep_total
(key: str)
src/skillevaluator/reporting/dashboard_data.py:622
↓ 1 callersFunctionsteps
(case_id: str, accuracy: float, step_judges: tuple[str, ...])
tests/test_harbor_metrics_truth.py:720
↓ 1 callersMethodstop
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:7
↓ 1 callersMethodstop
()
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodstop
(self, delete: bool)
src/skillevaluator/tier3/harbor/local_environment.py:645
↓ 1 callersMethodstring
(self, body: str, *, closed: bool)
src/skillevaluator/utils/structured_data.py:353
↓ 1 callersFunctionstring_open
(position: int)
src/skillevaluator/tier3/eval_core/checks.py:4997
↓ 1 callersFunctionstubborn_communication
()
tests/test_harbor_secure_docker_environment.py:7545
↓ 1 callersFunctionstubborn_communication
()
tests/test_harbor_local_mode.py:3700
↓ 1 callersFunctionsuccessful
(name: str)
tests/test_publication_evidence.py:303
↓ 1 callersFunctionsuccessful
(metric: str)
tests/tier3/test_judge_failure_artifacts.py:1314
↓ 1 callersFunctionsuccessful_result
(name: str, check: str)
tests/test_publication_identity.py:893
↓ 1 callersFunctionsummarize
(rows)
src/skillevaluator/tier3/grader_validation.py:375
↓ 1 callersFunctionsummarize
Return an action-item summary for a note.
tests/fixtures/skills/simple/scripts/summarize.py:12
↓ 1 callersFunctionsummarize_pinning
Counts of classified MCP pins plus ``pinned / (pinned + unpinned)``. ``ratio`` is ``None`` when no declaration runs a package (nothing to pin).
src/skillevaluator/plugin_mcp.py:180
↓ 1 callersMethodsummary
(self)
src/skillevaluator/reporting/presentation.py:66
↓ 1 callersFunctionsupport
(row: dict)
src/skillevaluator/reporting/plugin_static.py:366
↓ 1 callersFunctionsynthetic_trajectory_from_cline_cli
Build a trajectory from Cline CLI JSONL output (``cline.txt``). Cline CLI emits structured JSONL events with ``type``/``say``/``ask`` fields.
src/skillevaluator/tier3/eval_core/log_converters.py:270
↓ 1 callersFunctionsynthetic_trajectory_from_codex_json
Parse Codex ``exec --json`` ThreadEvent JSONL.
src/skillevaluator/tier3/eval_core/log_converters.py:814
↓ 1 callersFunctiontake_credentials
(start: int, end: int)
src/skillevaluator/tier3/eval_core/checks.py:4440
↓ 1 callersFunctiontargets
()
tests/tier3/test_mcp_proof.py:38
↓ 1 callersFunctiontask
(name: str, compose: str | None)
tests/test_harbor_environment_contract.py:263
↓ 1 callersFunctionte
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctiontest_every_workflow_uses_only_the_reviewed_npm_command
The reviewed lockfile install is the only npm command in workflows. npm accepts option abbreviations, aliases, and a changing command grammar.
tests/test_ci_workflows.py:854
↓ 1 callersFunctiontest_the_pinned_fern_cli_version_matches_the_fern_config
The Fern config, manifest, and lockfile describe one reviewed CLI tree. If they drift, docs may be validated or published with a different CLI or
tests/test_ci_workflows.py:867
↓ 1 callersMethodtext
:func:`diagnostic_text` of *value*, computed once per distinct text.
src/skillevaluator/skillspector_references.py:120
↓ 1 callersFunctionti
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctiontier3_options
Use the native service directly: combined validate requires Tier 1.
ci/run_tiers.py:225
↓ 1 callersFunctiontier3_result
(target: dict[str, str])
tests/test_publication_identity.py:898
↓ 1 callersFunctiontimed_out
(arm: str)
tests/tier3/test_token_usage.py:224
↓ 1 callersMethodto_dict
(self)
src/skillevaluator/plugin_dependencies.py:552
↓ 1 callersMethodto_dict
(self)
src/skillevaluator/harness/overlay.py:94
↓ 1 callersMethodto_dict
(self)
src/skillevaluator/tier3/plugin_native.py:2180
↓ 1 callersMethodto_dict
Convert to dictionary for JSON serialization.
src/skillevaluator/models/result.py:838
↓ 1 callersMethodto_dict
(self)
src/skillevaluator/models/quality.py:63
↓ 1 callersFunctiontoken_at
(limit: int)
tests/evaluation/test_tier3_feedback_chat.py:864
↓ 1 callersFunctiontolerated_baselines
Describe each agent kept with a failed baseline: ``{"name", "scored_attempts", "expected_attempts"}``.
src/skillevaluator/tier3/harbor/baseline_timeouts.py:324
↓ 1 callersFunctiontool
(name: str)
src/skillevaluator/plugin_dependency_audit.py:623
↓ 1 callersMethodtool_call
Read `tools.NAME(<object literal>)` or `tools.apply_patch(<string literal>)`.
src/skillevaluator/tier3/eval_core/codex_tool_call_normalizer.py:999
↓ 1 callersFunctiontop_argument_failures
The most frequent ``arguments.failures`` rows over per-trial signals, with exact counts. Every recorded row counts (a trial keeps at most 50), ke
src/skillevaluator/tier3/eval_core/plugin_signals.py:4117
↓ 1 callersFunctiontotal
(field: str)
src/skillevaluator/reporting/presentation.py:249
↓ 1 callersFunctiontracked_files
()
tests/ci/test_publisher_workflow.py:371
↓ 1 callersFunctiontrajectory
()
tests/tier3/test_plugin_evidence.py:54
↓ 1 callersFunctiontransfer
()
tests/test_harbor_secure_docker_environment.py:4208
↓ 1 callersFunctiontrial
(case, value, *, legacy=False)
tests/reporting/test_private_runtime_projection.py:343
↓ 1 callersFunctiontrial
(score: float, tokens: int)
tests/reporting/test_dashboard_app.py:41
↓ 1 callersFunctiontrial
(prompt: int, completion: int, seconds: float, cost: float)
tests/reporting/test_dashboard_data.py:24
↓ 1 callersFunctiontrial_quality_score
Return one trial's score on the same basis as the reported lift. The score is the mean of the trial's dimension scores. With *dimensions*, on
src/skillevaluator/tier3/harbor/stats.py:372
↓ 1 callersFunctiontrial_result
(trial_name: str)
src/skillevaluator/tier3/harbor/runner.py:4436
↓ 1 callersFunctiontrusted_blobs
Parse approved caller blobs exactly like the trusted feedback job.
src/skillevaluator/_ci_onboarding_all_tiers.py:844
↓ 1 callersFunctiontry_conversion
(expr: str)
src/skillevaluator/tier3/reference_skills/calculator/scripts/calc.py:63
↓ 1 callersFunctiontry_percentage
(expr: str)
src/skillevaluator/tier3/reference_skills/calculator/scripts/calc.py:81
↓ 1 callersFunctionts
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodtype
()
tests/test_harbor_claude_refusal_fallback.py:562
↓ 1 callersFunctionun
(t,e)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersFunctionunexplained_launch_errors
Return the launch errors that tolerated baselines do not explain, in order. Tolerance applies only when every launch error is a tolerated baselin
src/skillevaluator/tier3/harbor/baseline_timeouts.py:288
↓ 1 callersFunctionunique
(base: str)
src/skillevaluator/reporting/html.py:2032
↓ 1 callersFunctionunrelated
(rows)
ci/benchmark_publish.py:233
↓ 1 callersFunctionunscanned_skill_scan_incomplete_finding
LOW when the search for skills inside evaluation-output and snapshot folders stopped early. :func:`find_unscanned_plugin_skill_manifests` stopped
src/skillevaluator/plugin_member_paths.py:182
↓ 1 callersMethodunsupported_reason
(self, *, env_mode: str, task_source: str)
src/skillevaluator/tier3/plugin_native.py:1063
↓ 1 callersFunctionupdate
(r)
tests/test_tier3_history.py:129
↓ 1 callersMethodupdate
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateConfig
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateElements
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateElements
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateElements
(t,e,i,s)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateElements
(t,e,i,n)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateHoverStyle
(t,e,i)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
↓ 1 callersMethodupdateIndex
(t)
src/skillevaluator/reporting/assets/chart.umd.min.js:13
← previousnext →9,201–9,300 of 24,741, ranked by callers