inspect-ai 0.3.98__py3-none-any.whl → 0.3.100__py3-none-any.whl

inspect_ai/_view/www/log-schema.json

@@ -656,6 +656,17 @@
     "ContentAudio": {
       "description": "Audio content.",
       "properties": {
+        "internal": {
+          "anyOf": [
+            {
+              "$ref": "#/$defs/JsonValue"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null
+        },
         "type": {
           "const": "audio",
           "default": "audio",
@@ -676,6 +687,7 @@
         }
       },
       "required": [
+        "internal",
         "type",
         "audio",
         "format"
@@ -687,6 +699,17 @@
     "ContentImage": {
       "description": "Image content.",
       "properties": {
+        "internal": {
+          "anyOf": [
+            {
+              "$ref": "#/$defs/JsonValue"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null
+        },
         "type": {
           "const": "image",
           "default": "image",
@@ -709,6 +732,7 @@
         }
       },
       "required": [
+        "internal",
         "type",
         "image",
         "detail"
@@ -720,6 +744,17 @@
     "ContentReasoning": {
       "description": "Reasoning content.\n\nSee the specification for [thinking blocks](https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#understanding-thinking-blocks) for Claude models.",
       "properties": {
+        "internal": {
+          "anyOf": [
+            {
+              "$ref": "#/$defs/JsonValue"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null
+        },
         "type": {
           "const": "reasoning",
           "default": "reasoning",
@@ -749,6 +784,7 @@
         }
       },
       "required": [
+        "internal",
         "type",
         "reasoning",
         "signature",
@@ -761,6 +797,17 @@
     "ContentText": {
       "description": "Text content.",
       "properties": {
+        "internal": {
+          "anyOf": [
+            {
+              "$ref": "#/$defs/JsonValue"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null
+        },
         "type": {
           "const": "text",
           "default": "text",
@@ -785,6 +832,7 @@
         }
       },
       "required": [
+        "internal",
         "type",
         "text",
         "refusal"
@@ -796,6 +844,17 @@
     "ContentVideo": {
       "description": "Video content.",
       "properties": {
+        "internal": {
+          "anyOf": [
+            {
+              "$ref": "#/$defs/JsonValue"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null
+        },
         "type": {
           "const": "video",
           "default": "video",
@@ -817,6 +876,7 @@
         }
       },
       "required": [
+        "internal",
         "type",
         "video",
         "format"
@@ -1136,6 +1196,18 @@
           "default": null,
           "title": "Log Samples"
         },
+        "log_realtime": {
+          "anyOf": [
+            {
+              "type": "boolean"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null,
+          "title": "Log Realtime"
+        },
         "log_images": {
           "anyOf": [
             {
@@ -1205,6 +1277,7 @@
         "max_sandboxes",
         "sandbox_cleanup",
         "log_samples",
+        "log_realtime",
         "log_images",
         "log_buffer",
         "log_shared",
@@ -1502,7 +1575,8 @@
             "reasoning_tokens": null,
             "reasoning_summary": null,
             "reasoning_history": null,
-            "response_schema": null
+            "response_schema": null,
+            "extra_body": null
           }
         }
       },
@@ -1944,7 +2018,7 @@
       "additionalProperties": false
     },
     "EvalSampleLimit": {
-      "description": "Limit encontered by sample.",
+      "description": "Limit encountered by sample.",
       "properties": {
         "type": {
           "enum": [
@@ -1961,7 +2035,7 @@
         },
         "limit": {
           "title": "Limit",
-          "type": "integer"
+          "type": "number"
         }
       },
       "required": [
@@ -2277,6 +2351,10 @@
     "EvalSpec": {
       "description": "Eval target and configuration.",
       "properties": {
+        "eval_id": {
+          "title": "Eval Id",
+          "type": "string"
+        },
         "run_id": {
           "title": "Run Id",
           "type": "string"
@@ -2294,9 +2372,16 @@
           "type": "string"
         },
         "task_version": {
+          "anyOf": [
+            {
+              "type": "integer"
+            },
+            {
+              "type": "string"
+            }
+          ],
           "default": 0,
-          "title": "Task Version",
-          "type": "integer"
+          "title": "Task Version"
         },
         "task_file": {
           "anyOf": [
@@ -2500,6 +2585,7 @@
         }
       },
       "required": [
+        "eval_id",
         "run_id",
         "created",
         "task",
@@ -2897,6 +2983,19 @@
             }
           ],
           "default": null
+        },
+        "extra_body": {
+          "anyOf": [
+            {
+              "additionalProperties": true,
+              "type": "object"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null,
+          "title": "Extra Body"
         }
       },
       "title": "GenerateConfig",
@@ -2927,7 +3026,8 @@
         "reasoning_tokens",
         "reasoning_summary",
         "reasoning_history",
-        "response_schema"
+        "response_schema",
+        "extra_body"
       ],
       "additionalProperties": false
     },
@@ -4163,7 +4263,7 @@
         "limit": {
           "anyOf": [
             {
-              "type": "integer"
+              "type": "number"
             },
             {
               "type": "null"
@@ -5592,12 +5692,26 @@
         },
         "parameters": {
           "$ref": "#/$defs/ToolParams"
+        },
+        "options": {
+          "anyOf": [
+            {
+              "additionalProperties": true,
+              "type": "object"
+            },
+            {
+              "type": "null"
+            }
+          ],
+          "default": null,
+          "title": "Options"
         }
       },
       "required": [
         "name",
         "description",
-        "parameters"
+        "parameters",
+        "options"
       ],
       "title": "ToolInfo",
       "type": "object",

inspect_ai/_view/www/package.json

@@ -62,6 +62,8 @@
     "@codemirror/lint": "^6.8.5",
     "@codemirror/state": "^6.5.2",
     "@lezer/highlight": "^1.2.1",
+    "@mui/material": "^7.1.0",
+    "@mui/x-tree-view": "^8.3.0",
     "@popperjs/core": "^2.11.8",
     "ansi-output": "^0.0.9",
     "asciinema-player": "^3.9.0",
@@ -86,8 +88,10 @@
     "prismjs": "^1.30.0",
     "react": "^19.0.0",
     "react-dom": "^19.0.0",
+    "react-popper": "^2.3.0",
     "react-router-dom": "^7.5.3",
     "react-virtuoso": "^4.12.6",
-    "zustand": "^5.0.3"
+    "zustand": "^5.0.5",
+    "use-resize-observer": "^9.1.0"
   }
 }

inspect_ai/_view/www/src/@types/log.d.ts

@@ -7,11 +7,12 @@
 
 export type Version = number;
 export type Status = "started" | "success" | "cancelled" | "error";
+export type EvalId = string;
 export type RunId = string;
 export type Created = string;
 export type Task = string;
 export type TaskId = string;
-export type TaskVersion = number;
+export type TaskVersion = number | string;
 export type TaskFile = string | null;
 export type TaskRegistryName = string | null;
 export type Solver = string | null;
@@ -68,6 +69,9 @@ export type Anyof = JSONSchema[] | null;
 export type Required = string[] | null;
 export type Description1 = string | null;
 export type Strict = boolean | null;
+export type ExtraBody = {
+  [k: string]: unknown;
+} | null;
 export type ModelBaseUrl = string | null;
 export type ModelRoles = {
   [k: string]: EvalModelConfig;
@@ -99,6 +103,7 @@ export type MaxSubprocesses = number | null;
 export type MaxSandboxes = number | null;
 export type SandboxCleanup = boolean | null;
 export type LogSamples = boolean | null;
+export type LogRealtime = boolean | null;
 export type LogImages = boolean | null;
 export type LogBuffer = number | null;
 export type LogShared = number | null;
@@ -402,6 +407,9 @@ export type Description2 = string;
 export type Type11 = "object";
 export type Required1 = string[];
 export type Additionalproperties1 = boolean;
+export type Options3 = {
+  [k: string]: unknown;
+} | null;
 export type Tools1 = ToolInfo[];
 export type ToolChoice = ("auto" | "any" | "none") | ToolFunction;
 export type Name9 = string;
@@ -640,6 +648,7 @@ export interface EvalLog {
  * Eval target and configuration.
  */
 export interface EvalSpec {
+  eval_id: EvalId;
   run_id: RunId;
   created: Created;
   task: Task;
@@ -722,6 +731,7 @@ export interface GenerateConfig {
   reasoning_summary: ReasoningSummary;
   reasoning_history: ReasoningHistory;
   response_schema: ResponseSchema | null;
+  extra_body: ExtraBody;
 }
 /**
  * Schema for model response when using Structured Output.
@@ -786,6 +796,7 @@ export interface EvalConfig {
   max_sandboxes: MaxSandboxes;
   sandbox_cleanup: SandboxCleanup;
   log_samples: LogSamples;
+  log_realtime: LogRealtime;
   log_images: LogImages;
   log_buffer: LogBuffer;
   log_shared: LogShared;
@@ -888,6 +899,7 @@ export interface GenerateConfig1 {
   reasoning_summary: ReasoningSummary;
   reasoning_history: ReasoningHistory;
   response_schema: ResponseSchema | null;
+  extra_body: ExtraBody;
 }
 /**
  * Scoring results from evaluation.
@@ -998,6 +1010,7 @@ export interface ChatMessageSystem {
  * Text content.
  */
 export interface ContentText {
+  internal: unknown;
   type: Type3;
   text: Text;
   refusal: Refusal;
@@ -1008,6 +1021,7 @@ export interface ContentText {
  * See the specification for [thinking blocks](https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#understanding-thinking-blocks) for Claude models.
  */
 export interface ContentReasoning {
+  internal: unknown;
   type: Type4;
   reasoning: Reasoning;
   signature: Signature;
@@ -1017,6 +1031,7 @@ export interface ContentReasoning {
  * Image content.
  */
 export interface ContentImage {
+  internal: unknown;
   type: Type5;
   image: Image;
   detail: Detail;
@@ -1025,6 +1040,7 @@ export interface ContentImage {
  * Audio content.
  */
 export interface ContentAudio {
+  internal: unknown;
   type: Type6;
   audio: Audio;
   format: Format1;
@@ -1033,6 +1049,7 @@ export interface ContentAudio {
  * Video content.
  */
 export interface ContentVideo {
+  internal: unknown;
   type: Type7;
   video: Video;
   format: Format2;
@@ -1299,6 +1316,7 @@ export interface ToolInfo {
   name: Name8;
   description: Description2;
   parameters: ToolParams;
+  options: Options3;
 }
 /**
  * Description of tool parameters object in JSON Schema format.
@@ -1525,7 +1543,7 @@ export interface Attachments {
   [k: string]: string;
 }
 /**
- * Limit encontered by sample.
+ * Limit encountered by sample.
  */
 export interface EvalSampleLimit {
   type: Type16;

inspect_ai/_view/www/src/app/App.tsx

@@ -14,7 +14,7 @@ import "./App.css";
 
 import ClipboardJS from "clipboard";
 import { FC, useCallback, useEffect } from "react";
-import { RouterProvider, useParams } from "react-router-dom";
+import { RouterProvider } from "react-router-dom";
 import { ClientAPI, HostMessage } from "../client/api/types.ts";
 import { useStore } from "../state/store.ts";
 import { AppRouter } from "./routing/AppRouter.tsx";
@@ -32,7 +32,6 @@ export const App: FC<AppProps> = ({ api }) => {
 
   const logs = useStore((state) => state.logs.logs);
   const selectedLogFile = useStore((state) => state.logs.selectedLogFile);
-
   const loadedLogFile = useStore((state) => state.log.loadedLog);
   const selectedLogSummary = useStore((state) => state.log.selectedLogSummary);
 
@@ -46,15 +45,6 @@ export const App: FC<AppProps> = ({ api }) => {
   const loadLog = useStore((state) => state.logActions.loadLog);
   const pollLog = useStore((state) => state.logActions.pollLog);
 
-  const { sampleId } = useParams<{
-    logPath?: string;
-    tabId?: string;
-    sampleId?: string;
-    epoch?: string;
-    sampleTabId?: string;
-  }>();
-  const selectSample = useStore((state) => state.logActions.selectSample);
-
   // Load a specific log
   useEffect(() => {
     const loadSpecificLog = async () => {
@@ -66,10 +56,6 @@ export const App: FC<AppProps> = ({ api }) => {
           // Then load the log
           await loadLog(selectedLogFile);
 
-          if (!sampleId) {
-            selectSample(0);
-          }
-
           // Finally set loading to false
           setAppStatus({ loading: false, error: undefined });
         } catch (e) {

inspect_ai/_view/www/src/app/appearance/icons.ts

@@ -9,6 +9,7 @@ const loggingIcons: Record<string, string> = {
 };
 
 export const ApplicationIcons = {
+  agent: "bi bi-grid", // bi bi-x-diamond
   approve: "bi bi-shield",
   approvals: {
     approve: "bi bi-shield-check",
@@ -56,7 +57,7 @@ export const ApplicationIcons = {
   json: "bi bi-filetype-json",
   limits: {
     messages: "bi bi-chat-right-text",
-    custom: "bi bi-person-workspace",
+    custom: "bi bi-exclamation-triangle",
     operator: "bi bi-person-workspace",
     tokens: "bi bi-list",
     time: "bi bi-clock",
@@ -91,6 +92,7 @@ export const ApplicationIcons = {
   sandbox: "bi bi-box-seam",
   scorer: "bi bi-calculator",
   search: "bi bi-search",
+  sidebar: "bi bi-list",
   solvers: {
     default: "bi bi-arrow-return-right",
     generate: "bi bi-share",
@@ -106,5 +108,6 @@ export const ApplicationIcons = {
     open: "bi bi-caret-down-fill",
     closed: "bi bi-caret-right-fill",
   },
+  turns: "bi bi-chat-left-text", // bi bi-repeat
   usage: "bi bi-stopwatch",
 };

inspect_ai/_view/www/src/app/content/MetaDataGrid.tsx

@@ -7,6 +7,7 @@ interface MetadataGridProps {
   id?: string;
   className?: string | string[];
   style?: CSSProperties;
+  size?: "mini" | "small";
   entries: Record<string, unknown>;
   plain?: boolean;
 }
@@ -18,10 +19,13 @@ export const MetaDataGrid: FC<MetadataGridProps> = ({
   id,
   entries,
   className,
+  size,
   style,
   plain,
 }) => {
   const baseId = "metadata-grid";
+  const fontStyle =
+    size === "mini" ? "text-size-smallest" : "text-size-smaller";
 
   const entryEls = entryRecords(entries).map((entry, index) => {
     const id = `${baseId}-value-${index}`;
@@ -41,15 +45,29 @@ export const MetaDataGrid: FC<MetadataGridProps> = ({
             styles.cell,
             "text-style-label",
             "text-style-secondary",
-            "text-size-smaller",
+            fontStyle,
           )}
         >
-          {entry.name}
+          {entry?.name}
         </div>
-        <div
-          className={clsx(styles.value, `${baseId}-value`, "text-size-smaller")}
-        >
-          <RenderedContent id={id} entry={entry} />
+        <div className={clsx(styles.value, `${baseId}-value`, fontStyle)}>
+          {entry && (
+            <RenderedContent
+              id={id}
+              entry={entry}
+              renderObject={(obj: any) => {
+                return (
+                  <MetaDataGrid
+                    id={id}
+                    className={clsx(styles.nested)}
+                    entries={obj}
+                    size={size}
+                    plain={plain}
+                  />
+                );
+              }}
+            />
+          )}
         </div>
       </Fragment>
     );

inspect_ai/_view/www/src/app/content/MetadataGrid.module.css

@@ -9,8 +9,3 @@
   font-weight: 400;
   white-space: nowrap;
 }
-
-.value {
-  white-space: pre-wrap;
-  word-wrap: anywhere;
-}