Models: B

Baseline

What the candidate was compared with: recorded (the test set's recorded runs, with the versions that served them), or another version.

• kind: Literal['recorded']
• versions: list[kindgi.client.Version]

Baseline1

What the candidate was compared with: recorded (the test set's recorded runs, with the versions that served them), or another version.

• kind: Literal['version']
• agent_id: str (agentId on the wire)
• version: str
• via: Literal['explicit', 'live']
• live_scope: dict[str, Any] | None (liveScope on the wire) (optional)

Baseline2

An output's score: Σ yesWeight and Σ totalWeight over its judged items, and over those among the first k ranked items.

• yes_weight: float (yesWeight on the wire)
• total_weight: float (totalWeight on the wire)
• items: int
• judged_items: int (judgedItems on the wire)
• top_k: kindgi.client.TopK (topK on the wire)
• fresh: kindgi.client.Fresh | None (optional)

Binding

• root: str

BlobMeta

• blob_id: str (blobId on the wire)
• tenant_id: uuid.UUID (tenantId on the wire)
• name: str
• content_type: str (contentType on the wire)
• size: int
• hash: str
• tags: dict[str, str]
• owner_run_id: uuid.UUID | None (ownerRunId on the wire) (optional)
• project_id: str | None (projectId on the wire) (optional)
• created_by: str | None (createdBy on the wire) (optional)
• created_at: AwareDatetime (createdAt on the wire)
• bucket: str | None (optional)
• key: str | None (optional)

Block

A data block version: a prompt or settings, versioned like a tool (immutable versions, soft unregister). An agent version pins the block versions it uses when it is published. Belongs to one project and is authorized through it.

• id: str
• version: str
• kind: Union[Literal['prompt', 'settings'], str]
• description: str | None (optional)
• content: kindgi.client.PromptBlockContent | kindgi.client.SettingsBlockContent
• project_id: uuid.UUID (projectId on the wire)
• published_at: AwareDatetime (publishedAt on the wire)
• unregistered_at: AwareDatetime | None (unregisteredAt on the wire) (optional)

BlockCollectionPage

• data: list[kindgi.client.Block]
• next_cursor: str | None (nextCursor on the wire) (optional)
• has_more: bool (hasMore on the wire)

BlockRef

A settings block an agent reads, by id and semver range.

• id: str
• version: str

Budget

• max_cost_usd: float | None (maxCostUsd on the wire) (optional)
• max_latency_ms: int | None (maxLatencyMs on the wire) (optional)

Budget2

Default: $5 and 30 candidates.

• max_cost_usd: float | None (maxCostUsd on the wire) (optional)
• max_candidates: int | None (maxCandidates on the wire) (optional)

Budget3

Each pass's budget (default $5 and 30 candidates), never more than what's left of the month's cap.

• max_cost_usd: float | None (maxCostUsd on the wire) (optional)
• max_candidates: int | None (maxCandidates on the wire) (optional)

BuildJudgedSuiteBody

Name the agent (agentId) or the flow (flowId) whose judged runs to use.

• version: str
• project_id: str (projectId on the wire)
• agent_id: str | None (agentId on the wire) (optional)
• agent_version: str | None (agentVersion on the wire) (optional)
• flow_id: str | None (flowId on the wire) (optional)
• since: AwareDatetime | None (optional)
• until: AwareDatetime | None (optional)
• judge_class_ids: list[kindgi.client.JudgeClassId] | None (judgeClassIds on the wire) (optional)
• min_judgments: int | None (minJudgments on the wire) (optional)
• segments: list[kindgi.client.ScopeSegment] | None (optional)
• description: str | None (optional)

BuildJudgedSuiteResult

• suite_id: str (suiteId on the wire)
• version: str
• kind: Union[Literal['judged'], str]
• case_count: int (caseCount on the wire)
• truncated: bool