@prismshadow/penguin-server 0.1.4 → 0.1.5

package/dist/api/types.d.ts

@@ -116,6 +116,13 @@ interface ProjectCreateRequest {
 interface ProjectCreateResponse {
     project: ProjectSummary;
 }
+interface ProjectUpdateRequest {
+    /** New display name. The projectId itself is immutable — only this label can change. */
+    name: string;
+}
+interface ProjectUpdateResponse {
+    project: ProjectSummary;
+}
 interface MemberInfo {
     userId: string;
     role: ProjectRole;
@@ -511,7 +518,7 @@ interface MessagesResponse {
     live?: MessagesLiveTail;
 }
 /**
- * A single Prompt's input parts: text or image (data: / http(s) URL).
+ * A single Prompt's input parts: text, image (data: / http(s) URL), or an uploaded file.
  * Docs: /docs/server-api § "Session-Level Endpoints".
  */
 type TaskInputPart = {
@@ -520,6 +527,19 @@ type TaskInputPart = {
 } | {
     type: "image_url";
     imageUrl: string;
+}
+/**
+ * File attachment (the composer's "+" menu): `dataUrl` is a base64 `data:` URL of the
+ * file's bytes, capped at 10MB each (413 `file_too_large` beyond that; the request as a
+ * whole still has to fit the global 20MB body limit). The server writes it into the
+ * Session scratchpad under a sanitized name and appends an `[attached file: <path>]` line
+ * to the message text — the bytes never enter the conversation, the model opens the file
+ * by path. `fileName` is the original name (no path separators, no `..`).
+ */
+ | {
+    type: "file";
+    fileName: string;
+    dataUrl: string;
 };
 interface TaskCreateRequest {
     input: TaskInputPart[];
@@ -573,8 +593,16 @@ interface TaskCreateResponse {
  * task POST).
  */
 interface SteerRequest {
-    /** Non-empty message text (trimmed server-side). */
+    /** Message text (trimmed server-side); may be empty when `images` carries the message. */
     text: string;
+    /**
+     * Images sent with the steering message (`data:` or http(s) URLs, same rule as
+     * `TaskInputPart.image_url`): delivered as user image messages right behind the
+     * `[user_steering]` text. A model without vision receives them as scratchpad path lines
+     * instead, exactly as it would a Prompt's images. At least one of `text` / `images` must
+     * be non-empty.
+     */
+    images?: string[];
 }
 interface ApprovalDecisionRequest {
     decision: "allow" | "deny";
@@ -989,9 +1017,20 @@ interface UsageErrors {
     unexpected: number;
     /** The most frequent source · code (null when there are no errors). */
     topCode: UsageErrorCount | null;
-    /** Most recent N items (reverse chronological). */
+    /** Most recent N items (reverse chronological) — the first page; older ones come from `GET /usage/errors`. */
     recent: UsageErrorItem[];
 }
+/**
+ * GET /api/projects/:projectId/usage/errors — one page of the error detail table, newest
+ * first. The dashboard response above already carries the first page; this exists so
+ * "show me earlier ones" does not have to refetch the whole aggregate. It takes the same
+ * date/agent filter as the dashboard, so a page never widens what the summary counted.
+ */
+interface UsageErrorsPage {
+    items: UsageErrorItem[];
+    /** Filtered row count, so the caller knows when it has reached the end. */
+    total: number;
+}
 interface UsageResponse {
     summary: {
         today: UsageBucket;
@@ -1079,21 +1118,22 @@ interface AgentImportResponse {
 /** Raw result of a single run (a scoreboard per-case runs[] entry). */
 interface BenchmarkRunScore {
     score: number;
-    cost?: number;
-    durationMs?: number;
+    /** Run cost, or null when unavailable. */
+    cost: number | null;
+    durationMs: number;
     /** Id of the Session under test in this run (links to Trace). */
-    sessionId?: string;
+    sessionId: string;
 }
 interface BenchmarkCaseScore {
     case: string;
-    /** Per-case score = average of runs (equals that single run's score under the legacy single-run format). */
+    /** Model-written average of this Case's Run scores, on the fixed 0..100 scale. */
     score: number;
-    cost?: number;
-    durationMs?: number;
-    /** For legacy format compatibility: per-case single Session id (new format keeps it inside runs[]). */
-    sessionId?: string;
-    /** Raw results per run; unset under the legacy format (the server backfills one entry when parsing as a single run). */
-    runs?: BenchmarkRunScore[];
+    /** Model-written average of known Run costs; null when every Run cost is unknown. */
+    cost: number | null;
+    /** Model-written average of Run durations, rounded to an integer. */
+    durationMs: number;
+    /** Raw results per Run. */
+    runs: BenchmarkRunScore[];
 }
 interface BenchmarkEvaluation {
     /** Evaluation timestamp (ISO 8601). */
@@ -1103,15 +1143,19 @@ interface BenchmarkEvaluation {
     /** Evaluation summary body: how the score was derived, what optimizations were made to the Agent this round (required when generating, tolerated as unset when displaying). */
     summary?: string;
     /** Model actually used for this evaluation round (upstream id, paired with provider; the chart series is split by model). */
-    modelId?: string;
+    modelId: string;
     /** Provider group for `modelId`. */
-    provider?: string;
+    provider: string;
+    /** Thinking level read from the unchanged Target Agent configuration. */
+    thinkingLevel: string;
     /** Agent State version number under test. */
-    version?: number;
-    /** Total score (sum of per-case scores; max score defined by the scoring rubric). */
+    version: number;
+    /** Model-written average of Case scores, on the fixed 0..100 scale. */
     score: number;
-    cost?: number;
-    durationMs?: number;
+    /** Model-written average of known Case costs; null when every Case cost is unknown. */
+    cost: number | null;
+    /** Model-written average of Case durations, rounded to an integer. */
+    durationMs: number;
     cases: BenchmarkCaseScore[];
 }
 interface BenchmarkSummary {
@@ -1130,6 +1174,15 @@ interface BenchmarkSummary {
 interface BenchmarksResponse {
     benchmarks: BenchmarkSummary[];
 }
+/** Public Benchmark Case metadata. Rubric and Gold content are never included. */
+interface BenchmarkCaseSummary {
+    id: string;
+    /** First Markdown heading with an optional leading "Case N:" removed; falls back to id. */
+    title: string;
+}
+interface BenchmarkCasesResponse {
+    cases: BenchmarkCaseSummary[];
+}
 interface SkillMetadataItem {
     /** Skill directory name (the identity key for install / uninstall / Prompt addressing). */
     name: string;
@@ -1212,4 +1265,4 @@ interface UpdateRunResponse {
     needsRestart: boolean;
 }
 
-export type { AdminPasswordResetRequest, AdminUserCreateRequest, AdminUserCreateResponse, AdminUsersResponse, AgentCompactionConfigDto, AgentConfigDto, AgentConfigResponse, AgentConfigUpdateRequest, AgentCreateRequest, AgentCreateResponse, AgentImportRequest, AgentImportResponse, AgentModelConfigDto, AgentSkillsResponse, AgentSummary, AgentTraceDateGroup, AgentTraceFileRef, AgentTraceSessionGroup, AgentTracesResponse, AgentsResponse, ApprovalDecisionRequest, ApprovalMode, AuthLoginRequest, AuthResponse, BenchmarkCaseScore, BenchmarkEvaluation, BenchmarkRunScore, BenchmarkSummary, BenchmarksResponse, CredentialInfo, DirEntryInfo, DirListResponse, ErrorBody, FilesStatRequest, FilesStatResponse, GoalResponse, GoalServerEvent, GoalStateView, MeResponse, MemberAddRequest, MemberAddResponse, MemberInfo, MembersResponse, MessagesLiveTail, MessagesResponse, ModelInfo, ModelPricingDto, ModelRefDto, ModelTestRequest, ModelTestResponse, ModelUpdateEntry, ModelsResponse, ModelsUpdateRequest, PasswordChangeRequest, PrefsResponse, ProjectCreateRequest, ProjectCreateResponse, ProjectRole, ProjectSummary, ProjectsResponse, RequestSpan, RetryNowResponse, ScheduleItem, ScheduleServerEvent, ScheduleStatus, ScheduleUpsertRequest, SchedulesResponse, ServerEvent, SessionCategory, SessionCategoryCounts, SessionCreateRequest, SessionCreateResponse, SessionInfo, SessionPatchRequest, SessionResponse, SessionSource, SessionStatus, SessionTracesResponse, SessionsResponse, SkillGroupItem, SkillInstallRequest, SkillLibraryResponse, SkillMetadataItem, SteerRequest, TaskCreateRequest, TaskCreateResponse, TaskInputPart, ToolCallSpan, TraceAnalysisResponse, TraceEventsResponse, TraceFileInfo, TraceImportRequest, TraceImportResponse, TraceModelSegment, TraceTaskStats, TraceToolSpan, UiPrefs, UpdateCheckResponse, UpdateRunResponse, UsageAgentCount, UsageBucket, UsageErrorCount, UsageErrorItem, UsageErrors, UsageGroupBy, UsageGroupRow, UsageResponse, UsageSuccessRate, UsageTrendPoint, UsageTrendPointInTrace, UserInfo, VaultEntryInfo, VaultEntryUpdate, VaultResponse, VaultUpdateRequest, VersionResponse, WorkspaceFileEntry, WorkspaceFilesResponse };
+export type { AdminPasswordResetRequest, AdminUserCreateRequest, AdminUserCreateResponse, AdminUsersResponse, AgentCompactionConfigDto, AgentConfigDto, AgentConfigResponse, AgentConfigUpdateRequest, AgentCreateRequest, AgentCreateResponse, AgentImportRequest, AgentImportResponse, AgentModelConfigDto, AgentSkillsResponse, AgentSummary, AgentTraceDateGroup, AgentTraceFileRef, AgentTraceSessionGroup, AgentTracesResponse, AgentsResponse, ApprovalDecisionRequest, ApprovalMode, AuthLoginRequest, AuthResponse, BenchmarkCaseScore, BenchmarkCaseSummary, BenchmarkCasesResponse, BenchmarkEvaluation, BenchmarkRunScore, BenchmarkSummary, BenchmarksResponse, CredentialInfo, DirEntryInfo, DirListResponse, ErrorBody, FilesStatRequest, FilesStatResponse, GoalResponse, GoalServerEvent, GoalStateView, MeResponse, MemberAddRequest, MemberAddResponse, MemberInfo, MembersResponse, MessagesLiveTail, MessagesResponse, ModelInfo, ModelPricingDto, ModelRefDto, ModelTestRequest, ModelTestResponse, ModelUpdateEntry, ModelsResponse, ModelsUpdateRequest, PasswordChangeRequest, PrefsResponse, ProjectCreateRequest, ProjectCreateResponse, ProjectRole, ProjectSummary, ProjectUpdateRequest, ProjectUpdateResponse, ProjectsResponse, RequestSpan, RetryNowResponse, ScheduleItem, ScheduleServerEvent, ScheduleStatus, ScheduleUpsertRequest, SchedulesResponse, ServerEvent, SessionCategory, SessionCategoryCounts, SessionCreateRequest, SessionCreateResponse, SessionInfo, SessionPatchRequest, SessionResponse, SessionSource, SessionStatus, SessionTracesResponse, SessionsResponse, SkillGroupItem, SkillInstallRequest, SkillLibraryResponse, SkillMetadataItem, SteerRequest, TaskCreateRequest, TaskCreateResponse, TaskInputPart, ToolCallSpan, TraceAnalysisResponse, TraceEventsResponse, TraceFileInfo, TraceImportRequest, TraceImportResponse, TraceModelSegment, TraceTaskStats, TraceToolSpan, UiPrefs, UpdateCheckResponse, UpdateRunResponse, UsageAgentCount, UsageBucket, UsageErrorCount, UsageErrorItem, UsageErrors, UsageErrorsPage, UsageGroupBy, UsageGroupRow, UsageResponse, UsageSuccessRate, UsageTrendPoint, UsageTrendPointInTrace, UserInfo, VaultEntryInfo, VaultEntryUpdate, VaultResponse, VaultUpdateRequest, VersionResponse, WorkspaceFileEntry, WorkspaceFilesResponse };