fix: Write context/index.json so agents can look up tasks by phase

Agents were bulk-reading all context task files (39 files) because
filenames are opaque IDs and there was no way to find phase-relevant
tasks without reading every file. Now writeInputFiles generates a
context/index.json with tasksByPhase mapping phaseId to task metadata
(file, id, name, status). Prompt updated to direct agents to read
the index first.

apps/server/agent/file-io.test.ts

@@ -3,7 +3,7 @@
  */
 
 import { describe, it, expect, beforeEach, afterEach } from 'vitest';
-import { mkdirSync, writeFileSync, rmSync, existsSync } from 'node:fs';
+import { mkdirSync, writeFileSync, readFileSync, rmSync, existsSync } from 'node:fs';
 import { join } from 'node:path';
 import { tmpdir } from 'node:os';
 import { randomUUID } from 'crypto';
@@ -116,6 +116,34 @@ describe('writeInputFiles', () => {
     writeInputFiles({ agentWorkdir: testDir });
     expect(existsSync(join(testDir, '.cw', 'input'))).toBe(true);
   });
+
+  it('writes context/index.json grouping tasks by phaseId', () => {
+    writeInputFiles({
+      agentWorkdir: testDir,
+      tasks: [
+        { id: 't1', name: 'Task A', phaseId: 'ph1', status: 'pending', category: 'execute', type: 'auto', priority: 'medium' } as Task,
+        { id: 't2', name: 'Task B', phaseId: 'ph1', status: 'completed', category: 'execute', type: 'auto', priority: 'medium', summary: 'Done' } as Task,
+        { id: 't3', name: 'Task C', phaseId: 'ph2', status: 'pending', category: 'research', type: 'auto', priority: 'high' } as Task,
+      ],
+    });
+
+    const indexPath = join(testDir, '.cw', 'input', 'context', 'index.json');
+    expect(existsSync(indexPath)).toBe(true);
+    const index = JSON.parse(readFileSync(indexPath, 'utf-8'));
+    expect(index.tasksByPhase.ph1).toHaveLength(2);
+    expect(index.tasksByPhase.ph2).toHaveLength(1);
+    expect(index.tasksByPhase.ph1[0]).toEqual({
+      file: 'context/tasks/t1.md',
+      id: 't1',
+      name: 'Task A',
+      status: 'pending',
+    });
+  });
+
+  it('does not write context/index.json when no tasks', () => {
+    writeInputFiles({ agentWorkdir: testDir });
+    expect(existsSync(join(testDir, '.cw', 'input', 'context', 'index.json'))).toBe(false);
+  });
 });
 
 describe('readSummary', () => {

apps/server/agent/file-io.ts

@@ -262,6 +262,29 @@ export function writeInputFiles(options: WriteInputFilesOptions): void {
     }
   }
 
+  // Write context index — groups tasks by phaseId so agents can look up relevant files
+  // without bulk-reading every context file
+  if (options.tasks && options.tasks.length > 0) {
+    const tasksByPhase: Record<string, Array<{ file: string; id: string; name: string; status: string }>> = {};
+    for (const t of options.tasks) {
+      const phaseId = t.phaseId ?? '_unassigned';
+      if (!tasksByPhase[phaseId]) tasksByPhase[phaseId] = [];
+      tasksByPhase[phaseId].push({
+        file: `context/tasks/${t.id}.md`,
+        id: t.id,
+        name: t.name,
+        status: t.status,
+      });
+    }
+    const contextDir = join(inputDir, 'context');
+    mkdirSync(contextDir, { recursive: true });
+    writeFileSync(
+      join(contextDir, 'index.json'),
+      JSON.stringify({ tasksByPhase }, null, 2) + '\n',
+      'utf-8',
+    );
+  }
+
   // Write manifest listing exactly which files were created
   writeFileSync(
     join(inputDir, 'manifest.json'),

apps/server/agent/prompts/shared.ts

@@ -25,12 +25,14 @@ Read \`.cw/input/manifest.json\` first. It contains two arrays:
 - \`pages/\` — one per page; frontmatter: title, parentPageId, sortOrder; body: markdown
 
 **Context Files** (read-only, read on-demand)
+- \`context/index.json\` — **read this first** when you need context. Contains \`tasksByPhase\`: a map of phaseId → array of \`{ file, id, name, status }\`. Use it to find relevant task files without bulk-reading.
 - \`context/phases/\` — frontmatter: id, name, status, dependsOn; body: description
 - \`context/tasks/\` — frontmatter: id, name, phaseId, parentTaskId, category, type, priority, status, summary; body: description
   Completed tasks include a \`summary\` field with what the previous agent accomplished.
 
-Context files provide awareness of the broader initiative. There may be dozens — do NOT bulk-read them.
+Context files provide awareness of the broader initiative. There may be dozens — do NOT bulk-read them all.
-When you need to check a specific phase or task, read that one file. Do not duplicate or contradict context file content in your output.
+Use \`context/index.json\` to find which task files belong to a specific phase, then read only those.
+Do not duplicate or contradict context file content in your output.
 </input_files>`;
 
 export const ID_GENERATION = `