Skip to content

Tools

Cluster: Uncategorised | Type: component | MCP Tools: None

Overview

Public API

FriendlinessScorecard

Field Type Default
release str required
capability_delta int required
oversight_delta int required
blocked bool required
reason str ''

Methods:

to_dict() -> dict

HarnessScore(BaseModel)

Field Type Default
harness_id str required
mitigation_quality float 0.0
mitigation_quality_ci95_low float 0.0
denominator_n int 0
examples_learned int 0
skills_added int 0
verified_runs int 0
qualified_runs int 0
blocked_runs int 0

ReleaseScorecard(BaseModel)

Field Type Default
release_id str required
previous_release_id str ''
tasks_attempted int 0
tasks_succeeded int 0
first_pass_success_rate float 0.0
rollback_rate float 0.0
human_intervention_minutes float 0.0
average_component_completion_time float 0.0
components_under_10_minutes int 0
repeated_failure_modes list[str] Field(default_factory=list)
new_failure_modes list[str] Field(default_factory=list)
resolved_failure_modes list[str] Field(default_factory=list)
tests_added int 0
meta_evals_added int 0
docs_updated bool False
source_visualisation_updated bool False
routing_policy_changes int 0
cost_per_successful_task float 0.0
latency_per_successful_task float 0.0
codex_reviews_completed int 0
claude_design_reviews_completed int 0
human_approvals_required int 0
safety_blocks_triggered int 0
known_regressions list[str] Field(default_factory=list)
rollback_points list[str] Field(default_factory=list)
per_harness list[HarnessScore] Field(default_factory=list)
release_rule_phase str 'shadow'
release_rule_fired bool False
release_rule_reason str ''

ComponentScorecard

Field Type Default
component_id str required
tasks_verified int 0
qualified_drafts int 0
blocked_escalations int 0
recurring_failures list[str] field(default_factory=list)
average_cost float 0.0
average_latency float 0.0
rollback_count int 0
mitigation_quality dict[str, Any] field(default_factory=dict)

Functions

generate_scorecard(release: str, capability_delta: int = 0, oversight_delta: int = 0) -> FriendlinessScorecard

main(argv: list[str] | None = None) -> int

wilson_ci95_low(successes: int, total: int) -> float

evaluate_release_rule(card: ReleaseScorecard, previous: ReleaseScorecard | None = None, release_number: int = 1, cutover_approved: bool = False) -> tuple[bool, str, str]

generate_release_scorecard(release_id: str, previous_release_id: str = '', ledger: ReliabilityLedger | None = None, harness_ids: list[str] | None = None, previous: ReleaseScorecard | None = None, release_number: int = 1) -> ReleaseScorecard

write_scorecard(card: ReleaseScorecard, output_dir: str | Path) -> tuple[Path, Path]

main(argv: list[str] | None = None) -> int

generate_component_scorecard(component_id: str, rows: list[dict[str, Any]] | None = None) -> ComponentScorecard

scorecard_to_dict(card: ComponentScorecard) -> dict[str, Any]

write_scorecard(card: ComponentScorecard, output: str | Path) -> tuple[Path, Path]

main() -> int