refactor: extend media understanding

src/auto-reply/reply/commands-info.ts

@@ -68,6 +68,7 @@ export const handleStatusCommand: CommandHandler = async (params, allowTextComma
     resolveDefaultThinkingLevel: params.resolveDefaultThinkingLevel,
     isGroup: params.isGroup,
     defaultGroupActivation: params.defaultGroupActivation,
+    mediaDecisions: params.ctx.MediaUnderstandingDecisions,
   });
   return { shouldContinue: false, reply };
 };

src/auto-reply/reply/commands-status.ts

@@ -24,6 +24,7 @@ import type { ElevatedLevel, ReasoningLevel, ThinkLevel, VerboseLevel } from "..
 import type { ReplyPayload } from "../types.js";
 import type { CommandContext } from "./commands-types.js";
 import { getFollowupQueueDepth, resolveQueueSettings } from "./queue.js";
+import type { MediaUnderstandingDecision } from "../../media-understanding/types.js";
 
 function formatApiKeySnippet(apiKey: string): string {
   const compact = apiKey.replace(/\s+/g, "");
@@ -105,6 +106,7 @@ export async function buildStatusReply(params: {
   resolveDefaultThinkingLevel: () => Promise<ThinkLevel | undefined>;
   isGroup: boolean;
   defaultGroupActivation: () => "always" | "mention";
+  mediaDecisions?: MediaUnderstandingDecision[];
 }): Promise<ReplyPayload | undefined> {
   const {
     cfg,
@@ -200,6 +202,7 @@ export async function buildStatusReply(params: {
       dropPolicy: queueSettings.dropPolicy,
       showDetails: queueOverrides,
     },
+    mediaDecisions: params.mediaDecisions,
     includeTranscriptUsage: false,
   });
 

src/auto-reply/reply/get-reply-directives-apply.ts

@@ -188,6 +188,7 @@ export async function applyInlineDirectiveOverrides(params: {
         resolveDefaultThinkingLevel: async () => resolvedDefaultThinkLevel,
         isGroup,
         defaultGroupActivation: defaultActivation,
+        mediaDecisions: ctx.MediaUnderstandingDecisions,
       });
     }
     typing.cleanup();

src/auto-reply/reply/get-reply-inline-actions.ts

@@ -185,6 +185,7 @@ export async function handleInlineActions(params: {
       resolveDefaultThinkingLevel,
       isGroup,
       defaultGroupActivation: defaultActivation,
+      mediaDecisions: ctx.MediaUnderstandingDecisions,
     });
     await sendInlineReply(inlineStatusReply);
     directives = { ...directives, hasStatusDirective: false };

src/auto-reply/status.test.ts

@@ -90,6 +90,59 @@ describe("buildStatusMessage", () => {
     expect(text).toContain("elevated");
   });
 
+  it("includes media understanding decisions when present", () => {
+    const text = buildStatusMessage({
+      agent: { model: "anthropic/claude-opus-4-5" },
+      sessionEntry: { sessionId: "media", updatedAt: 0 },
+      sessionKey: "agent:main:main",
+      queue: { mode: "none" },
+      mediaDecisions: [
+        {
+          capability: "image",
+          outcome: "success",
+          attachments: [
+            {
+              attachmentIndex: 0,
+              attempts: [
+                {
+                  type: "provider",
+                  outcome: "success",
+                  provider: "openai",
+                  model: "gpt-5.2",
+                },
+              ],
+              chosen: {
+                type: "provider",
+                outcome: "success",
+                provider: "openai",
+                model: "gpt-5.2",
+              },
+            },
+          ],
+        },
+        {
+          capability: "audio",
+          outcome: "skipped",
+          attachments: [
+            {
+              attachmentIndex: 1,
+              attempts: [
+                {
+                  type: "provider",
+                  outcome: "skipped",
+                  reason: "maxBytes: too large",
+                },
+              ],
+            },
+          ],
+        },
+      ],
+    });
+
+    const normalized = normalizeTestText(text);
+    expect(normalized).toContain("Media: image ok (openai/gpt-5.2) · audio skipped (maxBytes)");
+  });
+
   it("does not show elevated label when session explicitly disables it", () => {
     const text = buildStatusMessage({
       agent: { model: "anthropic/claude-opus-4-5", elevatedDefault: "on" },

src/auto-reply/status.ts

@@ -24,6 +24,7 @@ import { VERSION } from "../version.js";
 import { listChatCommands, listChatCommandsForConfig } from "./commands-registry.js";
 import type { SkillCommandSpec } from "../agents/skills.js";
 import type { ElevatedLevel, ReasoningLevel, ThinkLevel, VerboseLevel } from "./thinking.js";
+import type { MediaUnderstandingDecision } from "../media-understanding/types.js";
 
 type AgentConfig = Partial<NonNullable<NonNullable<ClawdbotConfig["agents"]>["defaults"]>>;
 
@@ -52,6 +53,7 @@ type StatusArgs = {
   modelAuth?: string;
   usageLine?: string;
   queue?: QueueStatus;
+  mediaDecisions?: MediaUnderstandingDecision[];
   includeTranscriptUsage?: boolean;
   now?: number;
 };
@@ -167,6 +169,42 @@ const formatUsagePair = (input?: number | null, output?: number | null) => {
   return `🧮 Tokens: ${inputLabel} in / ${outputLabel} out`;
 };
 
+const formatMediaUnderstandingLine = (decisions?: MediaUnderstandingDecision[]) => {
+  if (!decisions || decisions.length === 0) return null;
+  const parts = decisions
+    .map((decision) => {
+      const count = decision.attachments.length;
+      const countLabel = count > 1 ? ` x${count}` : "";
+      if (decision.outcome === "success") {
+        const chosen = decision.attachments.find((entry) => entry.chosen)?.chosen;
+        const provider = chosen?.provider?.trim();
+        const model = chosen?.model?.trim();
+        const modelLabel = provider ? (model ? `${provider}/${model}` : provider) : null;
+        return `${decision.capability}${countLabel} ok${modelLabel ? ` (${modelLabel})` : ""}`;
+      }
+      if (decision.outcome === "no-attachment") {
+        return `${decision.capability} none`;
+      }
+      if (decision.outcome === "disabled") {
+        return `${decision.capability} off`;
+      }
+      if (decision.outcome === "scope-deny") {
+        return `${decision.capability} denied`;
+      }
+      if (decision.outcome === "skipped") {
+        const reason = decision.attachments
+          .flatMap((entry) => entry.attempts.map((attempt) => attempt.reason).filter(Boolean))
+          .find(Boolean);
+        const shortReason = reason ? reason.split(":")[0]?.trim() : undefined;
+        return `${decision.capability} skipped${shortReason ? ` (${shortReason})` : ""}`;
+      }
+      return null;
+    })
+    .filter(Boolean);
+  if (parts.length === 0) return null;
+  return `📎 Media: ${parts.join(" · ")}`;
+};
+
 export function buildStatusMessage(args: StatusArgs): string {
   const now = args.now ?? Date.now();
   const entry = args.sessionEntry;
@@ -320,12 +358,14 @@ export function buildStatusMessage(args: StatusArgs): string {
   const costLine = costLabel ? `💵 Cost: ${costLabel}` : null;
   const usageCostLine =
     usagePair && costLine ? `${usagePair} · ${costLine}` : (usagePair ?? costLine);
+  const mediaLine = formatMediaUnderstandingLine(args.mediaDecisions);
 
   return [
     versionLine,
     modelLine,
     usageCostLine,
     `📚 ${contextLine}`,
+    mediaLine,
     args.usageLine,
     `🧵 ${sessionLine}`,
     `⚙️ ${optionsLine}`,