mixdog 0.8.1 → 0.9.1

package/src/runtime/shared/tool-surface.mjs

@@ -1,10 +1,29 @@
+import { displayModelName as sharedDisplayModelName } from '../../ui/model-display.mjs';
+
 const DEFAULT_SUMMARY_MAX = 160;
+// Semantic cap for collapsed agent/task card one-liners (spawn, send, response).
+export const AGENT_SURFACE_BRIEF_MAX = 120;
 const STATUS_SEPARATOR = ' · ';
 
 export function stripToolPrefix(name) {
-  return String(name || 'tool')
-    .replace(/^mcp__.*__/, '')
-    .replace(/^functions\./, '');
+  const text = rawToolName(name);
+  const mcp = parseMcpToolName(text);
+  return mcp ? mcp.tool : text;
+}
+
+function rawToolName(name) {
+  return String(name || 'tool').replace(/^functions\./, '');
+}
+
+export function parseMcpToolName(name) {
+  const text = rawToolName(name);
+  const match = /^mcp__(.+?)__(.+)$/.exec(text);
+  if (!match) return null;
+  return { server: match[1], tool: match[2] };
+}
+
+export function isMcpToolName(name) {
+  return Boolean(parseMcpToolName(name));
 }
 
 export function normalizeToolName(name) {
@@ -61,6 +80,12 @@ function compactSlash(left, right) {
   return a && b ? `${a}/${b}` : a || b;
 }
 
+function mcpToolTarget(name, max = DEFAULT_SUMMARY_MAX) {
+  const mcp = parseMcpToolName(name);
+  if (!mcp) return '';
+  return truncateToolText(compactSlash(mcp.server, mcp.tool), max);
+}
+
 function quoted(value, max) {
   const text = truncateToolText(value || '', max);
   return text ? `"${text}"` : '';
@@ -73,6 +98,52 @@ function firstText(...values) {
   return '';
 }
 
+function splitToolSearchSelection(value) {
+  if (Array.isArray(value)) return value.map((item) => String(item || '').trim()).filter(Boolean);
+  return String(value || '')
+    .split(/[,\s]+/)
+    .map((item) => item.trim())
+    .filter(Boolean);
+}
+
+function toolSearchTargetKind(value) {
+  const lower = String(value || '').trim().toLowerCase();
+  if (!lower) return '';
+  if (lower.startsWith('mcp__') || lower.includes('_mcp_') || lower.includes('mcp')) return 'MCP';
+  if (lower === 'skill' || lower.startsWith('skill:') || lower.startsWith('skill_') || lower.startsWith('skills_') || lower.includes('skill')) return 'Skills';
+  return 'Tools';
+}
+
+function orderedToolSearchKinds(kinds) {
+  const set = new Set((kinds || []).filter(Boolean));
+  const order = ['Tools', 'MCP', 'Skills'];
+  const out = order.filter((kind) => set.has(kind));
+  return out.length ? out : ['Tools'];
+}
+
+function inferToolSearchKinds(args = {}) {
+  const selected = splitToolSearchSelection(args.select);
+  if (selected.length) return orderedToolSearchKinds(selected.map(toolSearchTargetKind));
+
+  const query = firstText(args.query, args.q, args.text).toLowerCase();
+  const kinds = [];
+  if (/\bmcp\b|mcp__|mcp[-_\s]?server/.test(query)) kinds.push('MCP');
+  if (/\bskills?\b|skill_|use[-_\s]?skill/.test(query)) kinds.push('Skills');
+  return orderedToolSearchKinds(kinds);
+}
+
+function toolSearchDisplayLabel(args = {}) {
+  return `Load ${inferToolSearchKinds(args).join('/')}`;
+}
+
+function displayToolSearchTarget(value) {
+  const text = String(value || '').trim();
+  if (/^skill:/i.test(text)) return text.slice('skill:'.length);
+  const mcp = /^mcp__(.*?)__(.+)$/.exec(text);
+  if (mcp) return `${mcp[1]}.${mcp[2]}`;
+  return stripToolPrefix(text);
+}
+
 function titleizeToolName(name) {
   return stripToolPrefix(name)
     .split(/[_\s-]+/)
@@ -89,7 +160,6 @@ function titleizeToolName(name) {
 
 const AGENT_DISPLAY_NAMES = new Map([
   ['explore', 'Explore'],
-  ['web-researcher', 'Web Researcher'],
   ['maintainer', 'Maintainer'],
   ['worker', 'Worker'],
   ['heavy-worker', 'Heavy Worker'],
@@ -121,41 +191,22 @@ function displayAgentName(value) {
   return AGENT_DISPLAY_NAMES.get(key) || titleizeDisplayName(text);
 }
 
-function displayModelName(model) {
-  const text = String(model || '').trim();
-  if (!text) return '';
-  const raw = text.includes('/') ? (text.split('/').filter(Boolean).at(-1) || text) : text;
-  const lower = raw.toLowerCase();
-  const newClaude = /^claude-(opus|sonnet|haiku)-(\d+)(?:-(\d+))?/i.exec(lower);
-  if (newClaude) {
-    const family = titleizeDisplayName(newClaude[1]);
-    const minor = newClaude[3] && newClaude[3].length <= 2 ? `.${newClaude[3]}` : '';
-    return `${family} ${newClaude[2]}${minor}`;
-  }
-  const oldClaude = /^claude-(\d+)(?:-(\d+))?-(opus|sonnet|haiku)(?:-|$)/i.exec(lower);
-  if (oldClaude) {
-    const family = titleizeDisplayName(oldClaude[3]);
-    return `${family} ${oldClaude[1]}${oldClaude[2] ? `.${oldClaude[2]}` : ''}`;
-  }
-  if (lower.startsWith('gpt-')) {
-    return raw
-      .split('-')
-      .map((part, index) => (index === 0 ? part.toUpperCase() : part.charAt(0).toUpperCase() + part.slice(1)))
-      .join('-');
-  }
-  if (lower.startsWith('grok-')) {
-    return raw
-      .split('-')
-      .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
-      .join(' ');
-  }
-  return raw;
+export function displayModelName(model, provider, displayHint) {
+  const text = String(model ?? '').trim();
+  const modelId = text
+    ? (text.includes('/') ? (text.split('/').filter(Boolean).at(-1) || text) : text)
+    : '';
+  const shown = sharedDisplayModelName(modelId, provider, displayHint);
+  return shown || modelId;
 }
 
 function bridgeAgentModelSummary(args) {
+  const provider = firstText(args.provider, args.providerId, args.provider_id);
+  const modelId = firstText(args.model);
+  const displayHint = firstText(args.modelDisplay, args.model_display, args.displayModel);
   return compactParts([
     displayAgentName(firstText(args.agent, args.role, args.name, args.subagent_type)),
-    displayModelName(firstText(args.modelDisplay, args.model_display, args.displayModel, args.model)),
+    displayModelName(modelId, provider, displayHint),
   ]);
 }
 
@@ -186,6 +237,36 @@ function summarizePatch(patch, basePath) {
   return text ? 'patch' : '';
 }
 
+function collectionCount(...values) {
+  for (const value of values) {
+    if (Array.isArray(value)) {
+      const count = value.filter((item) => item != null && String(item).trim()).length;
+      return count || value.length || 0;
+    }
+    if (value && typeof value === 'object') return 1;
+    if (value != null && String(value).trim()) return 1;
+  }
+  return 0;
+}
+
+function formatCountedUnit(count, singular, pluralText = `${singular}s`) {
+  const n = Math.max(0, Number(count || 0));
+  return n > 0 ? `${n} ${pluralize(n, singular, pluralText)}` : '';
+}
+
+function patchFileCount(args = {}) {
+  const patchText = String(args.patch ?? '');
+  if (patchText) {
+    const files = new Set();
+    for (const line of patchText.split('\n')) {
+      const match = /^\*\*\*\s+(?:Update|Add|Delete) File:\s+(.+)\s*$/.exec(line);
+      if (match) files.add(match[1].trim());
+    }
+    if (files.size > 0) return files.size;
+  }
+  return collectionCount(args.path, args.file, args.file_path, args.base_path);
+}
+
 function codeGraphLabel(args) {
   const mode = String(args.mode || args.action || '').toLowerCase();
   if (mode === 'prewarm' || mode === 'index' || mode === 'build' || mode === 'refresh') return 'Setup';
@@ -201,6 +282,7 @@ function codeGraphSummary(args, max) {
 }
 
 export function displayToolName(name, args = {}) {
+  if (isMcpToolName(name)) return 'MCP';
   const normalized = normalizeToolName(name);
   switch (normalized) {
     case 'read':
@@ -209,24 +291,32 @@ export function displayToolName(name, args = {}) {
       return 'Read';
     case 'apply_patch': {
       const parsed = parseToolArgs(args);
+      if (parsed && parsed.dry_run === true) return 'Check';
       return parsed && parsed.old_string === '' ? 'Create' : 'Update';
     }
     case 'shell':
+    case 'bash':
     case 'bash_session':
     case 'shell_command':
-    case 'task':
-    case 'trigger_schedule':
+    case 'job_wait':
       return 'Run';
+    case 'task':
+      return 'Task';
     case 'grep':
+    case 'find':
     case 'glob':
-    case 'tool_search':
+    case 'list':
+    case 'ls':
       return 'Search';
+    case 'trigger_schedule':
+      return 'Schedule';
+    case 'tool_search':
+      return toolSearchDisplayLabel(parseToolArgs(args));
     case 'search':
     case 'search_query':
     case 'image_query':
     case 'web_search':
     case 'web_search_call':
-    case 'firecrawl_search':
       return 'Web Search';
     case 'explore':
       return 'Explore';
@@ -282,9 +372,20 @@ export function displayToolName(name, args = {}) {
 export function summarizeToolArgs(name, args, { max = DEFAULT_SUMMARY_MAX } = {}) {
   const a = parseToolArgs(args);
   if (!a || typeof a !== 'object') return '';
-  switch (normalizeToolName(name)) {
+  const normalized = normalizeToolName(name);
+  const mcpTarget = mcpToolTarget(name, max);
+  if (mcpTarget) {
+    return compactParts([
+      mcpTarget,
+      truncateToolText(firstText(a.query, a.q, a.text, a.prompt, a.path, a.uri, a.name, a.id, a.action), Math.min(max, 80)),
+    ]);
+  }
+  switch (normalized) {
     case 'read':
       if (!a.path && !a.file_path) return '';
+      if (Array.isArray(a.path) || Array.isArray(a.file_path)) {
+        return formatCountedUnit(collectionCount(a.path, a.file_path), 'file');
+      }
       return compactParts([
         displayToolPath(a.path ?? a.file_path),
         a.pages ? `pages ${a.pages}` : summarizeLineWindow(a),
@@ -294,19 +395,27 @@ export function summarizeToolArgs(name, args, { max = DEFAULT_SUMMARY_MAX } = {}
     case 'apply_patch':
       return summarizePatch(a.patch, a.base_path);
     case 'shell':
+    case 'bash':
     case 'bash_session':
     case 'shell_command':
+    case 'job_wait':
       return truncateCommand(a.description || a.command || a.cmd || '', max);
     case 'task':
       return compactParts([a.action || a.type || 'task', a.task_id || '']);
     case 'list':
     case 'ls':
+      if (Array.isArray(a.path) || Array.isArray(a.dir) || Array.isArray(a.cwd)) {
+        return formatCountedUnit(collectionCount(a.path, a.dir, a.cwd), 'directory', 'directories');
+      }
       return compactParts([
         displayToolPath(a.path ?? a.dir ?? a.cwd ?? ''),
         a.head_limit || a.limit ? `${a.head_limit ?? a.limit} entries` : '',
       ]);
     case 'grep':
       if (!a.pattern && !a.query) return '';
+      if (Array.isArray(a.pattern) || Array.isArray(a.query)) {
+        return formatCountedUnit(collectionCount(a.pattern, a.query), 'pattern');
+      }
       return compactParts([
         `pattern: ${quoted(a.pattern ?? a.query, max)}`,
         a.path ? `path: ${displayToolPath(a.path)}` : '',
@@ -314,23 +423,47 @@ export function summarizeToolArgs(name, args, { max = DEFAULT_SUMMARY_MAX } = {}
       ]);
     case 'glob':
       if (!a.pattern && !a.glob) return '';
+      if (Array.isArray(a.pattern) || Array.isArray(a.glob)) {
+        return formatCountedUnit(collectionCount(a.pattern, a.glob), 'glob');
+      }
       return compactParts([
         `pattern: ${quoted(a.pattern ?? a.glob, max)}`,
         a.path ? `path: ${displayToolPath(a.path)}` : '',
       ]);
+    case 'find':
+      if (!a.query && !a.fuzzy) return '';
+      if (Array.isArray(a.query) || Array.isArray(a.fuzzy)) {
+        return formatCountedUnit(collectionCount(a.query, a.fuzzy), 'query', 'queries');
+      }
+      return compactParts([
+        quoted(a.query ?? a.fuzzy, max),
+        a.path ? `path: ${displayToolPath(a.path)}` : '',
+      ]);
     case 'search':
     case 'search_query':
     case 'image_query':
     case 'web_search':
     case 'web_search_call':
-    case 'firecrawl_search':
+      if (Array.isArray(a.query) || Array.isArray(a.keywords)) {
+        return formatCountedUnit(collectionCount(a.query, a.keywords), 'query', 'queries');
+      }
       return quoted(a.query || a.keywords || '', max);
     case 'explore':
+      if (Array.isArray(a.query) || Array.isArray(a.prompt) || Array.isArray(a.task) || Array.isArray(a.goal)) {
+        return formatCountedUnit(collectionCount(a.query, a.prompt, a.task, a.goal), 'query', 'queries');
+      }
       return truncateSingleLine(firstText(a.query, a.prompt, a.task, a.goal, a.path), Math.min(max, 80));
     case 'tool_search':
-      return quoted(firstText(a.query, a.q, a.text), max);
+      {
+        const selected = splitToolSearchSelection(a.select);
+        if (selected.length) return truncateToolText(selected.map(displayToolSearchTarget).join(', '), max);
+        return quoted(firstText(a.query, a.q, a.text), max);
+      }
     case 'web_fetch':
     case 'fetch':
+      if (Array.isArray(a.url) || Array.isArray(a.uri)) {
+        return formatCountedUnit(collectionCount(a.url, a.uri), 'URL', 'URLs');
+      }
       return truncateToolText(a.url || a.uri || '', max);
     case 'download_attachment':
       return displayToolPath(a.filename || a.name || a.url || '');
@@ -396,7 +529,7 @@ export function summarizeToolArgs(name, args, { max = DEFAULT_SUMMARY_MAX } = {}
     case 'skill_view':
     case 'skills_list':
     case 'use_skill':
-      return truncateToolText(firstText(a.name, a.skill, a.skill_name), max);
+      return truncateToolText(firstText(a.name, a.skill, a.skill_name, a.query, a.q, normalized === 'skills_list' ? 'all skills' : ''), max);
     default: {
       const primary = firstText(a.name, a.skill, a.query, a.title, a.path, a.file, a.target, a.id, a.action);
       if (primary) return truncateToolText(primary, Math.min(max, 80));
@@ -477,10 +610,13 @@ function parseLineDelta(delta) {
 
 function formatLineDelta(totals) {
   if (!totals?.seen) return '';
+  const added = Number(totals.added) || 0;
+  const removed = Number(totals.removed) || 0;
+  if (added === 0 && removed === 0) return '';
   const parts = [];
-  if (totals.added) parts.push(`+${totals.added} Lines`);
-  if (totals.removed) parts.push(`-${totals.removed} Lines`);
-  return parts.join(STATUS_SEPARATOR) || '±0 Lines';
+  if (added > 0) parts.push(`+${added} ${pluralize(added, 'line')}`);
+  if (removed > 0) parts.push(`-${removed} ${pluralize(removed, 'line')}`);
+  return parts.join(STATUS_SEPARATOR);
 }
 
 function parseUpdateSummary(text) {
@@ -510,6 +646,10 @@ function looksLineOriented(text) {
 }
 
 function summarizeUpdateResult(text, args) {
+  // A dry_run patch validates without writing, so the collapsed detail must not
+  // claim a real mutation ("Updated/Created/Deleted foo.js"). Map every action
+  // to "Checked" wording, matching the dry-run header (Checking/Checked).
+  const isDryRun = parseToolArgs(args)?.dry_run === true;
   const changed = [];
   for (const line of String(text ?? '').split('\n')) {
     const ok = /^\s*OK\s+(modify|add|delete|create)\s+(.+?)\s*$/i.exec(line);
@@ -526,8 +666,10 @@ function summarizeUpdateResult(text, args) {
   }
   if (changed.length === 1) {
     const item = changed[0];
-    const action = item.action === 'delete' ? 'Deleted' : item.action === 'add' || item.action === 'create' || item.action === 'created' ? 'Created' : 'Updated';
-    return compactParts([`${action} ${displayToolPath(item.path)}`, item.delta]);
+    const action = isDryRun
+      ? 'Checked'
+      : item.action === 'delete' ? 'Deleted' : item.action === 'add' || item.action === 'create' || item.action === 'created' ? 'Created' : 'Updated';
+    return compactParts([`${action} ${displayToolPath(item.path)}`, formatLineDelta(parseLineDelta(item.delta))]);
   }
   if (changed.length > 1) {
     const totals = changed.reduce((acc, item) => {
@@ -537,12 +679,12 @@ function summarizeUpdateResult(text, args) {
       acc.seen = acc.seen || delta.seen;
       return acc;
     }, { added: 0, removed: 0, seen: false });
-    return compactParts([`Updated ${changed.length} Files`, formatLineDelta(totals)]);
+    return compactParts([`${isDryRun ? 'Checked' : 'Updated'} ${changed.length} Files`, formatLineDelta(totals)]);
   }
 
   const parsedArgs = parseToolArgs(args);
   const target = parsedArgs.path ?? parsedArgs.file ?? parsedArgs.file_path ?? '';
-  if (target) return `Updated ${displayToolPath(target)}`;
+  if (target) return `${isDryRun ? 'Checked' : 'Updated'} ${displayToolPath(target)}`;
   return null;
 }
 
@@ -552,21 +694,66 @@ function textBetweenTag(text, tag) {
   return match ? match[1].trim() : '';
 }
 
+// Strip the most common inline-markdown markers so a one-line card summary
+// reads as plain prose ("**not clean**" → "not clean", "`x`" → "x"). Block
+// markers (#, >, -, 1.) at line start are dropped too. Whitespace is collapsed
+// by the caller's truncateSingleLine.
+function stripInlineMarkdown(value) {
+  return String(value ?? '')
+    .replace(/^\s{0,3}(?:#{1,6}\s+|>\s+|[-*+]\s+|\d+\.\s+)/, '')
+    .replace(/\*\*([^*]+)\*\*/g, '$1')
+    .replace(/__([^_]+)__/g, '$1')
+    .replace(/(?<!\*)\*(?!\*)([^*]+)\*(?!\*)/g, '$1')
+    .replace(/`([^`]+)`/g, '$1')
+    .trim();
+}
+
 function firstAgentResultLine(text) {
   const finalAnswer = textBetweenTag(text, 'final-answer') || textBetweenTag(text, 'result');
   const raw = finalAnswer || text;
   for (const line of String(raw ?? '').split('\n')) {
     const trimmed = line.trim();
     if (!trimmed) continue;
-    if (/^bridge result\b/i.test(trimmed)) continue;
+    if (/^agent result\b/i.test(trimmed)) continue;
     if (/^<\/?(?:final-answer|task-notification|task-id|tool-use-id|output-file|result|status|summary|usage|total_tokens|tool_uses|duration_ms|worktree|worktreePath|worktreeBranch)[^>]*>$/i.test(trimmed)) continue;
-    if (/^(?:bridge job|bridge task|status|type|target|role|agent|preset|model|effort|fast|limits|session|job|task-id|task_id|notification|queueDepth):\s*/i.test(trimmed)) continue;
+    if (/^(?:agent task|status|type|target|role|agent|preset|model|effort|fast|limits|session|task-id|task_id|notification|queueDepth):\s*/i.test(trimmed)) continue;
     if (/^\[[a-z-]+:\s*[^\]]*\]$/i.test(trimmed)) continue;
-    return truncateSingleLine(trimmed, 120);
+    return truncateSingleLine(trimmed, AGENT_SURFACE_BRIEF_MAX);
   }
   return '';
 }
 
+function summarizeGenericResult(text) {
+  const trimmed = String(text ?? '').trim();
+  if (!trimmed) return null;
+  if (/^(?:undefined|null)$/i.test(trimmed)) return null;
+
+  if (/^[\[{]/.test(trimmed)) {
+    try {
+      const parsed = JSON.parse(trimmed);
+      if (Array.isArray(parsed)) return `${parsed.length} ${pluralize(parsed.length, 'item')}`;
+      if (parsed && typeof parsed === 'object') {
+        const status = firstText(parsed.status, parsed.state, parsed.result, parsed.message);
+        if (status) return truncateSingleLine(titleStatus(status), 120);
+        if (parsed.cwd) return truncateSingleLine(parsed.cwd, 120);
+        if (typeof parsed.ok === 'boolean') return parsed.ok ? 'Ok' : 'Failed';
+        for (const key of ['items', 'results', 'resources', 'templates', 'providers', 'schedules', 'channels', 'tools']) {
+          if (Array.isArray(parsed[key])) return `${parsed[key].length} ${pluralize(parsed[key].length, (key.slice(0, -1) || 'item').toLowerCase())}`;
+        }
+      }
+    } catch {
+      return null;
+    }
+  }
+
+  const line = firstAgentResultLine(text) || trimmed.split('\n').map((item) => item.trim()).find(Boolean) || '';
+  if (!line || line === '{' || line === '[') return null;
+  if (/^(ok|done|success|saved|sent|updated|reloaded|connected|enabled|disabled|active|inactive)$/i.test(line)) {
+    return titleStatus(line);
+  }
+  return truncateSingleLine(line, 120);
+}
+
 /**
  * Derive a short semantic one-liner for a completed tool call using only the
  * tool name, parsed args, and the raw result text. Returns null when nothing
@@ -576,6 +763,8 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
   if (isError) return null;
   const text = String(resultText ?? '');
   const trimmed = text.trim();
+  if (/^(?:undefined|null)$/i.test(trimmed)) return null;
+  if (isMcpToolName(name)) return trimmed ? firstAgentResultLine(text) || null : null;
   const normalized = normalizeToolName(name);
 
   switch (normalized) {
@@ -585,7 +774,7 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
       if (/^\[image:/i.test(trimmed)) return 'Image';
       if (!trimmed) return null;
       const n = text.split('\n').length;
-      return `${n} ${pluralize(n, 'Line')}`;
+      return `${n} ${pluralize(n, 'line')}`;
     }
     case 'apply_patch': {
       const updateSummary = summarizeUpdateResult(text, args);
@@ -613,17 +802,32 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
       if (!trimmed || !looksLineOriented(text)) return null;
       const n = countNonEmptyLines(text);
       if (n === 0) return null;
-      return `${n} ${pluralize(n, 'Match', 'Matches')}`;
+      return `${n} ${pluralize(n, 'match', 'matches')}`;
     }
     case 'glob': {
       if (!trimmed || !looksLineOriented(text)) return null;
       const n = countNonEmptyLines(text);
       if (n === 0) return null;
-      return `${n} ${pluralize(n, 'File')}`;
+      return `${n} ${pluralize(n, 'file')}`;
+    }
+    case 'find': {
+      if (!trimmed || !looksLineOriented(text)) return null;
+      const n = countNonEmptyLines(text);
+      if (n === 0) return null;
+      return `${n} ${pluralize(n, 'candidate')}`;
+    }
+    case 'list':
+    case 'ls': {
+      if (!trimmed || !looksLineOriented(text)) return null;
+      const n = countNonEmptyLines(text);
+      if (n === 0) return null;
+      return `${n} ${pluralize(n, 'entry', 'entries')}`;
     }
     case 'shell':
+    case 'bash':
     case 'bash_session':
-    case 'shell_command': {
+    case 'shell_command':
+    case 'job_wait': {
       if (!trimmed) return '(No Output)';
       const job = /^\[(?:task_id|job):\s*([^\]]+)\]/mi.exec(text);
       const status = /^\[status:\s*([^\]]+)\]/mi.exec(text);
@@ -635,11 +839,12 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
           exit ? `Exit ${exit[1]}` : '',
         ]);
       }
-      return null;
+      const firstLine = trimmed.split('\n').map((line) => line.trim()).find(Boolean) || trimmed;
+      return truncateSingleLine(firstLine, 120);
     }
     case 'code_graph': {
       const match = /(\d+)\s+(references|definitions|symbols|callers|callees|results|matches)/i.exec(text);
-      if (match) return `${match[1]} ${titleWord(match[2])}`;
+      if (match) return `${match[1]} ${String(match[2]).toLowerCase()}`;
       return null;
     }
     case 'web_fetch':
@@ -654,35 +859,94 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
       if (status) return `HTTP ${status[1]}`;
       return null;
     }
+    case 'download_attachment':
+      return summarizeGenericResult(text);
     case 'search': {
       const match = /(\d+)\s+results?/i.exec(text);
       if (match) {
         const n = Number(match[1]);
-        return `${n} ${pluralize(n, 'Result')}`;
+        return `${n} ${pluralize(n, 'result')}`;
       }
       return null;
     }
     case 'search_query':
     case 'image_query':
     case 'web_search':
-    case 'web_search_call':
-    case 'firecrawl_search': {
+    case 'web_search_call': {
       const match = /(\d+)\s+results?/i.exec(text);
       if (match) {
         const n = Number(match[1]);
-        return `${n} ${pluralize(n, 'Result')}`;
+        return `${n} ${pluralize(n, 'result')}`;
       }
       return trimmed ? firstAgentResultLine(text) || null : null;
     }
     case 'explore': {
       return trimmed ? firstAgentResultLine(text) || null : null;
     }
-    case 'bridge':
+    case 'recall':
+    case 'search_memories':
+    case 'memory':
+    case 'remember':
+    case 'save_memory':
+    case 'update_memory':
+    case 'reply':
+    case 'react':
+    case 'edit_message':
+    case 'activate_channel_bridge':
+    case 'inject_command':
+    case 'request_user_input':
+    case 'update_plan':
+    case 'diagnostics':
+    case 'open_config':
+    case 'provider_status':
+    case 'channel_status':
+    case 'schedule_status':
+    case 'schedule_control':
+    case 'trigger_schedule':
+    case 'reload_config':
+    case 'cwd':
+    case 'list_mcp_resources':
+    case 'list_mcp_resource_templates':
+      return summarizeGenericResult(text);
+    case 'skill':
+    case 'skill_execute':
+    case 'skill_view':
+    case 'skills_list':
+    case 'use_skill': {
+      const parsedArgs = parseToolArgs(args);
+      const target = firstText(parsedArgs.name, parsedArgs.skill, parsedArgs.skill_name);
+      if (normalized === 'skills_list') {
+        const count = /(\d+)\s+skills?/i.exec(text);
+        if (count) return `${Number(count[1]) || count[1]} ${pluralize(Number(count[1]) || 0, 'skill')}`;
+        const lines = countNonEmptyLines(text);
+        return lines > 0 ? `${lines} ${pluralize(lines, 'skill')}` : null;
+      }
+      if (target) {
+        const verb = normalized === 'skill' || normalized === 'skill_view' ? 'Loaded' : 'Used';
+        return `${verb} ${truncateToolText(target, 80)}`;
+      }
+      return trimmed ? firstAgentResultLine(text) || null : null;
+    }
     case 'agent':
     case 'task': {
+      // Status-check (list/status) envelopes start with "agents: N" / "tasks: M"
+      // (or "(no agents or tasks)"). Collapse them to a tight count summary
+      // instead of leaking the raw "agents: 3 …" worker dump into the card.
+      if (/^\(no agents or tasks\)$/im.test(text)) return 'No agents or tasks';
+      const agentsCount = /^agents:\s*(\d+)/im.exec(text);
+      const tasksCount = /^tasks:\s*(\d+)/im.exec(text);
+      if (agentsCount || tasksCount) {
+        const a = agentsCount ? Number(agentsCount[1]) : 0;
+        const t = tasksCount ? Number(tasksCount[1]) : 0;
+        const parts = [];
+        if (agentsCount) parts.push(`${a} ${pluralize(a, 'agent')}`);
+        if (tasksCount && t > 0) parts.push(`${t} ${pluralize(t, 'task')}`);
+        return compactParts(parts) || 'No agents or tasks';
+      }
       const answerLine = firstAgentResultLine(text);
-      if (answerLine) return answerLine;
-      const job = /^bridge job:\s*(job_[^\s]+)/mi.exec(text);
+      // Agent/task result cards show only a one-liner; full report via ctrl+o.
+      if (answerLine) return truncateSingleLine(stripInlineMarkdown(answerLine), AGENT_SURFACE_BRIEF_MAX);
+      const task = /^agent task:\s*(\S+)/mi.exec(text);
       const status = /^status:\s*([^\s(]+)/mi.exec(text);
       const role = /^role:\s*(.+)$/mi.exec(text);
       const preset = /^preset:\s*(.+)$/mi.exec(text);
@@ -694,6 +958,7 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
       ]);
       if (agentModel) return agentModel;
       const parts = [
+        task ? task[1] : '',
         role ? role[1] : '',
         preset ? preset[1] : '',
         model ? model[1] : '',
@@ -701,7 +966,7 @@ export function summarizeToolResult(name, args, resultText, isError = false) {
         limits ? limits[1] : '',
       ].filter(Boolean);
       if (parts.length) return compactParts(parts);
-      if (job) return status ? `${job[1]} ${titleStatus(status[1])}` : job[1];
+      if (task) return status ? `${task[1]} ${titleStatus(status[1])}` : task[1];
       return null;
     }
     default:
@@ -713,6 +978,36 @@ export function isExplorerSurface(label) {
   return label === 'Read' || label === 'Search';
 }
 
+function truncateAgentSurfaceBrief(value, max = AGENT_SURFACE_BRIEF_MAX) {
+  const text = String(value ?? '').replace(/\s+/g, ' ').trim();
+  if (!text) return '';
+  if (text.length <= max) return text;
+  return `${text.slice(0, Math.max(1, max - 1))}\u2026`;
+}
+
+// Tight one-liner for an agent/task card: prefer the inbound prompt/message for
+// spawn/send, or a stripped first line of the result when a response landed.
+export function summarizeAgentSurfaceBrief(name, args, resultText, { isError = false, isResponse = false } = {}) {
+  const a = parseToolArgs(args);
+  const action = String(a?.type || a?.action || '').toLowerCase();
+  const text = String(resultText ?? '').trim();
+  if (isResponse && text) {
+    const fromResult = summarizeToolResult(name, args, text, isError);
+    if (fromResult) return truncateAgentSurfaceBrief(stripInlineMarkdown(fromResult));
+    const line = firstAgentResultLine(text);
+    if (line) return truncateAgentSurfaceBrief(stripInlineMarkdown(line));
+  }
+  const outbound = firstText(a?.prompt, a?.message);
+  if (outbound && (action === 'spawn' || action === 'send' || !action)) {
+    return truncateAgentSurfaceBrief(outbound);
+  }
+  if ((action === 'spawn' || action === 'send') && text && !isResponse) {
+    const fromResult = summarizeToolResult(name, args, text, isError);
+    if (fromResult) return truncateAgentSurfaceBrief(stripInlineMarkdown(fromResult));
+  }
+  return '';
+}
+
 export function isMemorySurface(label) {
   return label === 'Memory';
 }
@@ -720,8 +1015,8 @@ export function isMemorySurface(label) {
 // ── Aggregate tool-card classification & formatting ──────────────
 
 export const CATEGORY_ORDER = [
-  'Read', 'Search', 'Web Research', 'Memory', 'Explore',
-  'Patch', 'Shell', 'Agent', 'Channel', 'Setup', 'Other',
+  'Read', 'Search', 'Load', 'MCP', 'Skill', 'Web Research', 'Memory', 'Explore',
+  'Patch', 'Shell', 'Agent', 'Task', 'Schedule', 'Channel', 'Setup', 'Other',
 ];
 
 const TOOL_CATEGORY = new Map([
@@ -729,16 +1024,16 @@ const TOOL_CATEGORY = new Map([
   ['view_image', 'Read'],
   ['read_mcp_resource', 'Read'],
   ['grep', 'Search'],
+  ['find', 'Search'],
   ['glob', 'Search'],
   ['list', 'Search'],
   ['ls', 'Search'],
-  ['tool_search', 'Search'],
+  ['tool_search', 'Load'],
   ['search', 'Web Research'],
   ['web_search', 'Web Research'],
   ['search_query', 'Web Research'],
   ['image_query', 'Web Research'],
   ['web_search_call', 'Web Research'],
-  ['firecrawl_search', 'Web Research'],
   ['web_fetch', 'Web Research'],
   ['fetch', 'Web Research'],
   ['download_attachment', 'Web Research'],
@@ -756,8 +1051,7 @@ const TOOL_CATEGORY = new Map([
   ['shell_command', 'Shell'],
   ['bash_session', 'Shell'],
   ['job_wait', 'Shell'],
-  ['task', 'Agent'],
-  ['bridge', 'Agent'],
+  ['task', 'Task'],
   ['agent', 'Agent'],
   ['reply', 'Channel'],
   ['react', 'Channel'],
@@ -776,16 +1070,17 @@ const TOOL_CATEGORY = new Map([
   ['cwd', 'Setup'],
   ['request_user_input', 'Setup'],
   ['update_plan', 'Setup'],
-  ['trigger_schedule', 'Setup'],
-  ['skill', 'Setup'],
-  ['skill_execute', 'Setup'],
-  ['skill_view', 'Setup'],
-  ['skills_list', 'Setup'],
-  ['use_skill', 'Setup'],
+  ['trigger_schedule', 'Schedule'],
+  ['skill', 'Skill'],
+  ['skill_execute', 'Skill'],
+  ['skill_view', 'Skill'],
+  ['skills_list', 'Skill'],
+  ['use_skill', 'Skill'],
 ]);
 
 /** Return the aggregate category for a tool name + args. */
 export function classifyToolCategory(name, args = {}) {
+  if (isMcpToolName(name)) return 'MCP';
   const normalized = normalizeToolName(name);
   if (normalized === 'code_graph') {
     const mode = String(args.mode || args.action || '').toLowerCase();
@@ -796,16 +1091,21 @@ export function classifyToolCategory(name, args = {}) {
 }
 
 const CATEGORY_COPY = new Map([
-  ['Read', { active: 'Reading', done: 'Read', noun: 'item' }],
-  ['Search', { active: 'Searching', done: 'Searched', noun: 'item' }],
-  ['Web Research', { active: 'Researching', done: 'Researched', noun: 'web item' }],
+  ['Read', { active: 'Reading', done: 'Read', noun: 'file' }],
+  ['Search', { active: 'Searching', done: 'Searched', noun: 'file' }],
+  ['Load', { active: 'Loading', done: 'Loaded', noun: 'tool' }],
+  ['MCP', { active: 'Using', done: 'Used', noun: 'MCP tool' }],
+  ['Skill', { active: 'Loading', done: 'Loaded', noun: 'skill' }],
+  ['Web Research', { active: 'Researching', done: 'Researched', noun: 'query', pluralNoun: 'queries' }],
   ['Memory', { active: 'Checking', done: 'Checked', noun: 'memory item' }],
-  ['Explore', { active: 'Exploring', done: 'Explored', noun: 'item' }],
-  ['Patch', { active: 'Editing', done: 'Edited', noun: 'item' }],
+  ['Explore', { active: 'Exploring', done: 'Explored', noun: 'query', pluralNoun: 'queries' }],
+  ['Patch', { active: 'Editing', done: 'Edited', noun: 'file' }],
   ['Shell', { active: 'Running', done: 'Ran', noun: 'command' }],
   ['Agent', { active: 'Calling', done: 'Called', noun: 'agent' }],
+  ['Task', { active: 'Checking', done: 'Checked', noun: 'task' }],
+  ['Schedule', { active: 'Running', done: 'Ran', noun: 'schedule' }],
   ['Channel', { active: 'Sending', done: 'Sent', noun: 'message' }],
-  ['Setup', { active: 'Setting Up', done: 'Set Up', noun: 'item' }],
+  ['Setup', { active: 'Setting up', done: 'Set up', noun: 'item' }],
   ['Other', { active: 'Calling', done: 'Called', noun: 'tool' }],
 ]);
 
@@ -823,17 +1123,241 @@ function categoryNoun(category, count) {
   return pluralize(count, copy.noun, copy.pluralNoun || `${copy.noun}s`);
 }
 
+function categoryCopy(category) {
+  return CATEGORY_COPY.get(category) || CATEGORY_COPY.get('Other') || { active: 'Calling', done: 'Called', noun: 'tool' };
+}
+
+function unitDescriptor(category, overrides = {}) {
+  const copy = categoryCopy(category);
+  return {
+    category,
+    active: overrides.active || copy.active,
+    done: overrides.done || copy.done,
+    noun: overrides.noun || copy.noun || 'item',
+    pluralNoun: overrides.pluralNoun || copy.pluralNoun || `${overrides.noun || copy.noun || 'item'}s`,
+    count: Math.max(1, Number(overrides.count || 1)),
+  };
+}
+
+function queryCount(args, ...keys) {
+  return collectionCount(...keys.map((key) => args?.[key]));
+}
+
+export function toolWorkUnit(name, args = {}, category = '') {
+  const a = parseToolArgs(args);
+  const normalized = normalizeToolName(name);
+  const cat = category || classifyToolCategory(name, a);
+  if (isMcpToolName(name)) {
+    return unitDescriptor('MCP', { count: queryCount(a, 'query', 'q', 'text', 'prompt', 'path', 'uri', 'name', 'id', 'action') || 1 });
+  }
+  switch (normalized) {
+    case 'read':
+      return unitDescriptor('Read', { count: queryCount(a, 'path', 'paths', 'file_path', 'file', 'files') || 1, noun: 'file' });
+    case 'view_image':
+      return unitDescriptor('Read', { count: queryCount(a, 'path', 'file_path', 'file') || 1, noun: 'image' });
+    case 'read_mcp_resource':
+      return unitDescriptor('Read', { count: queryCount(a, 'uri', 'uris') || 1, noun: 'resource' });
+    case 'apply_patch': {
+      const creating = a.old_string === '';
+      // A dry_run patch validates the diff WITHOUT writing any file, so the
+      // header must not claim "Editing/Edited" (which made a pure validation
+      // look like a real edit). Surface it as "Checking/Checked" instead.
+      if (a.dry_run === true) {
+        return unitDescriptor('Patch', {
+          count: patchFileCount(a) || 1,
+          active: 'Checking',
+          done: 'Checked',
+          noun: 'file',
+        });
+      }
+      return unitDescriptor('Patch', {
+        count: patchFileCount(a) || 1,
+        active: creating ? 'Creating' : 'Editing',
+        done: creating ? 'Created' : 'Edited',
+        noun: 'file',
+      });
+    }
+    case 'grep':
+      return unitDescriptor('Search', { count: queryCount(a, 'pattern', 'patterns', 'query') || 1, active: 'Searching', done: 'Searched', noun: 'pattern' });
+    case 'glob':
+      return unitDescriptor('Search', { count: queryCount(a, 'pattern', 'patterns', 'glob', 'globs') || 1, active: 'Finding', done: 'Found', noun: 'glob' });
+    case 'find':
+      return unitDescriptor('Search', { count: queryCount(a, 'query', 'queries', 'fuzzy') || 1, active: 'Finding', done: 'Found', noun: 'query', pluralNoun: 'queries' });
+    case 'list':
+    case 'ls':
+      return unitDescriptor('Search', { count: queryCount(a, 'path', 'paths', 'dir', 'dirs', 'cwd') || 1, active: 'Listing', done: 'Listed', noun: 'directory', pluralNoun: 'directories' });
+    case 'tool_search': {
+      const selected = splitToolSearchSelection(a.select);
+      if (selected.length) return unitDescriptor('Load', { count: selected.length, noun: 'tool' });
+      return unitDescriptor('Load', { count: queryCount(a, 'query', 'q', 'text') || 1, noun: 'query', pluralNoun: 'queries' });
+    }
+    case 'search':
+    case 'search_query':
+    case 'image_query':
+    case 'web_search':
+    case 'web_search_call':
+      return unitDescriptor('Web Research', { count: queryCount(a, 'query', 'queries', 'keywords') || 1, noun: 'query', pluralNoun: 'queries' });
+    case 'web_fetch':
+    case 'fetch':
+      return unitDescriptor('Web Research', { count: queryCount(a, 'url', 'urls', 'uri', 'uris') || 1, active: 'Fetching', done: 'Fetched', noun: 'URL', pluralNoun: 'URLs' });
+    case 'download_attachment':
+      return unitDescriptor('Web Research', { count: queryCount(a, 'url', 'urls', 'filename', 'name') || 1, active: 'Downloading', done: 'Downloaded', noun: 'attachment' });
+    case 'recall':
+    case 'recall_memory':
+    case 'search_memories':
+      return unitDescriptor('Memory', { count: queryCount(a, 'query', 'queries', 'text', 'input') || 1, noun: 'memory item', pluralNoun: 'memory items' });
+    case 'remember':
+    case 'save_memory':
+    case 'update_memory':
+    case 'memory':
+      return unitDescriptor('Memory', { count: queryCount(a, 'entries', 'items', 'memories', 'query', 'text', 'value') || 1, active: 'Writing', done: 'Wrote', noun: 'memory item' });
+    case 'explore':
+      return unitDescriptor('Explore', { count: queryCount(a, 'query', 'queries', 'prompt', 'task', 'goal') || 1, noun: 'query', pluralNoun: 'queries' });
+    case 'shell':
+    case 'bash':
+    case 'bash_session':
+    case 'shell_command':
+    case 'job_wait':
+      return unitDescriptor('Shell', { count: queryCount(a, 'command', 'commands', 'cmd') || 1, noun: 'command' });
+    case 'agent':
+    case 'bridge':
+      return unitDescriptor('Agent', { count: queryCount(a, 'agents', 'roles', 'role', 'tag', 'task_id', 'sessionId') || 1, noun: 'agent' });
+    case 'task':
+      return unitDescriptor('Task', { count: queryCount(a, 'task_id', 'task_ids', 'id', 'ids') || 1, noun: 'task' });
+    case 'skill':
+    case 'skill_execute':
+    case 'skill_view':
+    case 'skills_list':
+    case 'use_skill':
+      return unitDescriptor('Skill', { count: queryCount(a, 'name', 'skill', 'skill_name', 'query', 'q') || 1, noun: 'skill' });
+    case 'reply':
+      return unitDescriptor('Channel', { count: queryCount(a, 'messages', 'messageId', 'text') || 1, noun: 'message' });
+    case 'react':
+      return unitDescriptor('Channel', { count: queryCount(a, 'emoji', 'messageId') || 1, active: 'Reacting', done: 'Reacted', noun: 'reaction' });
+    case 'edit_message':
+      return unitDescriptor('Channel', { count: queryCount(a, 'messageId', 'text') || 1, active: 'Editing', done: 'Edited', noun: 'message' });
+    case 'activate_channel_bridge':
+      return unitDescriptor('Channel', { active: 'Toggling', done: 'Toggled', noun: 'channel bridge' });
+    case 'inject_command':
+      return unitDescriptor('Channel', { count: queryCount(a, 'command', 'text', 'name') || 1, active: 'Injecting', done: 'Injected', noun: 'command' });
+    case 'trigger_schedule':
+      return unitDescriptor('Schedule', { count: queryCount(a, 'name', 'id') || 1, noun: 'schedule' });
+    case 'code_graph': {
+      const mode = String(a.mode || a.action || '').toLowerCase();
+      const searching = mode === 'search' || mode === 'find_symbol' || mode === 'references' || mode === 'callers' || mode === 'callees';
+      return unitDescriptor(searching ? 'Search' : 'Read', {
+        count: queryCount(a, 'symbols', 'symbol', 'query', 'file', 'path') || 1,
+        active: searching ? 'Mapping' : 'Reading',
+        done: searching ? 'Mapped' : 'Read',
+        noun: searching ? 'symbol' : 'file',
+      });
+    }
+    case 'request_user_input':
+      return unitDescriptor('Setup', { active: 'Asking', done: 'Asked', noun: 'user' });
+    case 'update_plan':
+      return unitDescriptor('Setup', { active: 'Updating', done: 'Updated', noun: 'plan' });
+    case 'diagnostics':
+      return unitDescriptor('Setup', { active: 'Checking', done: 'Checked', noun: 'diagnostic' });
+    case 'open_config':
+      return unitDescriptor('Setup', { active: 'Opening', done: 'Opened', noun: 'config' });
+    case 'provider_status':
+      return unitDescriptor('Setup', { active: 'Checking', done: 'Checked', noun: 'provider' });
+    case 'channel_status':
+      return unitDescriptor('Setup', { active: 'Checking', done: 'Checked', noun: 'channel' });
+    case 'schedule_status':
+      return unitDescriptor('Setup', { active: 'Checking', done: 'Checked', noun: 'schedule' });
+    case 'schedule_control':
+      return unitDescriptor('Setup', { active: 'Updating', done: 'Updated', noun: 'schedule' });
+    case 'reload_config':
+      return unitDescriptor('Setup', { active: 'Reloading', done: 'Reloaded', noun: 'config' });
+    case 'list_mcp_resources':
+      return unitDescriptor('Setup', { active: 'Listing', done: 'Listed', noun: 'MCP resource' });
+    case 'list_mcp_resource_templates':
+      return unitDescriptor('Setup', { active: 'Listing', done: 'Listed', noun: 'MCP resource template' });
+    case 'cwd': {
+      const action = String(a.action || a.type || '').toLowerCase();
+      return action === 'set'
+        ? unitDescriptor('Setup', { active: 'Setting', done: 'Set', noun: 'working directory', pluralNoun: 'working directories' })
+        : unitDescriptor('Setup', { active: 'Checking', done: 'Checked', noun: 'working directory', pluralNoun: 'working directories' });
+    }
+    default:
+      return unitDescriptor(cat, { count: queryCount(a, 'items', 'targets', 'query', 'path', 'name', 'id', 'action') || 1 });
+  }
+}
+
+function lifecycleVerb(unit, pending, { stableVerbWidth = false } = {}) {
+  const active = String(unit.active || '');
+  const done = String(unit.done || '');
+  const verb = pending ? active : done;
+  if (!stableVerbWidth) return verb;
+  return verb.padEnd(Math.max(active.length, done.length), ' ');
+}
+
+export function formatToolActionHeader(name, args = {}, { pending = false, count = 1, category = '', stableVerbWidth = false } = {}) {
+  const unit = toolWorkUnit(name, args, category);
+  const n = Math.max(1, Number(unit.count || count || 1));
+  const verb = lifecycleVerb(unit, pending, { stableVerbWidth });
+  return `${verb} ${n} ${pluralize(n, unit.noun, unit.pluralNoun)}`;
+}
+
+export function aggregateToolCategoryEntry(name, args = {}, category = '') {
+  const cat = category || classifyToolCategory(name, args);
+  const unit = toolWorkUnit(name, args, cat);
+  const key = [cat, unit.active, unit.done, unit.noun, unit.pluralNoun].join('|');
+  return {
+    key,
+    category: cat,
+    active: unit.active,
+    done: unit.done,
+    noun: unit.noun,
+    pluralNoun: unit.pluralNoun,
+    count: Math.max(1, Number(unit.count || 1)),
+  };
+}
+
+function aggregateCount(value) {
+  if (value && typeof value === 'object') return Math.max(0, Number(value.count || 0));
+  return Math.max(0, Number(value || 0));
+}
+
+function aggregateDescriptor(key, value) {
+  if (value && typeof value === 'object') {
+    const category = value.category || String(key || '').split('|')[0] || 'Other';
+    const copy = categoryCopy(category);
+    const noun = value.noun || copy.noun || 'item';
+    return {
+      category,
+      active: value.active || copy.active,
+      done: value.done || copy.done,
+      noun,
+      pluralNoun: value.pluralNoun || copy.pluralNoun || `${noun}s`,
+      count: aggregateCount(value),
+    };
+  }
+  const category = String(key || '');
+  const copy = categoryCopy(category);
+  const noun = copy.noun || 'item';
+  return {
+    category,
+    active: copy.active,
+    done: copy.done,
+    noun,
+    pluralNoun: copy.pluralNoun || `${noun}s`,
+    count: aggregateCount(value),
+  };
+}
+
 /**
  * Build a comma-separated header from per-category counts.
  * e.g. "Read 6 items, Searched 5 items, Called 1 agent"
  */
-export function formatAggregateHeader(categories, { pending = false, order = null } = {}) {
+export function formatAggregateHeader(categories, { pending = false, order = null, stableVerbWidth = false } = {}) {
   const categoryKeys = Object.keys(categories || {});
   const preferred = Array.isArray(order) && order.length ? order : categoryKeys;
   const seen = new Set();
   const ordered = [];
   const add = (cat) => {
-    if (!cat || seen.has(cat) || (categories[cat] || 0) <= 0) return;
+    if (!cat || seen.has(cat) || aggregateCount(categories[cat]) <= 0) return;
     seen.add(cat);
     ordered.push(cat);
   };
@@ -843,10 +1367,9 @@ export function formatAggregateHeader(categories, { pending = false, order = nul
 
   return ordered
     .map((cat) => {
-      const count = Number(categories[cat] || 0);
-      const label = pending ? activeCategoryLabel(cat) : doneCategoryLabel(cat);
-      if (pending) return label;
-      return `${label} ${count} ${categoryNoun(cat, count)}`;
+      const item = aggregateDescriptor(cat, categories[cat]);
+      const label = lifecycleVerb(item, pending, { stableVerbWidth });
+      return `${label} ${item.count} ${pluralize(item.count, item.noun, item.pluralNoun)}`;
     })
     .join(', ');
 }
@@ -882,13 +1405,13 @@ export function formatAggregateDetail(summaries) {
 
     let match = /^(?:Read\s+)?(\d+)\s+lines?$/i.exec(text);
     if (match) {
-      const metric = addMetric('read_lines', { count: 0, render: (m) => `${m.count} ${pluralize(m.count, 'Line')}` });
+      const metric = addMetric('read_lines', { count: 0, render: (m) => `${m.count} ${pluralize(m.count, 'line')}` });
       metric.count += Number(match[1]);
       continue;
     }
 
     if (/^(?:Read\s+)?image$/i.test(text)) {
-      const metric = addMetric('read_images', { count: 0, render: (m) => `${m.count} ${pluralize(m.count, 'Image')}` });
+      const metric = addMetric('read_images', { count: 0, render: (m) => `${m.count} ${pluralize(m.count, 'image')}` });
       metric.count += 1;
       continue;
     }
@@ -899,7 +1422,7 @@ export function formatAggregateDetail(summaries) {
       const singular = nounRaw.endsWith('ies') ? `${nounRaw.slice(0, -3)}y` : nounRaw.endsWith('s') ? nounRaw.slice(0, -1) : nounRaw;
       const plural = nounRaw.endsWith('s') ? nounRaw : `${nounRaw}s`;
       const key = `found_${plural}`;
-      const metric = addMetric(key, { count: 0, singular, plural, render: (m) => `${m.count} ${pluralize(m.count, titleWord(m.singular), titleWord(m.plural))}` });
+      const metric = addMetric(key, { count: 0, singular, plural, render: (m) => `${m.count} ${pluralize(m.count, m.singular, m.plural)}` });
       metric.count += Number(match[1]);
       continue;
     }
@@ -922,10 +1445,16 @@ export function formatAggregateDetail(summaries) {
         removed: 0,
         seen: false,
         render: (m) => {
+          // The aggregate header already carries the action + file count
+          // (e.g. "Edited 2 files"), so the detail row shows only the merged
+          // line delta. Fall back to the action + file/count summary only when
+          // there is no +/- delta to show (e.g. pure create/delete).
+          const delta = formatLineDelta(m);
+          if (delta) return delta;
           const count = m.fileCount + m.files.size;
           const action = m.actions.size === 1 ? [...m.actions][0] : 'Updated';
-          const target = count === 1 && m.fileCount === 0 ? [...m.files][0] : `${count} ${pluralize(count, 'File')}`;
-          return compactParts([`${action} ${target}`, formatLineDelta(m)]);
+          const target = count === 1 && m.fileCount === 0 ? [...m.files][0] : `${count} ${pluralize(count, 'file')}`;
+          return `${action} ${target}`;
         },
       });
       if (update.file) metric.files.add(update.file);