Merge remote-tracking branch 'origin/main' into brandon/cli-input-fixes

# Conflicts:
#	cli/src/state/chat-store.ts

Author: brandonkachen

.agents/__tests__/context-pruner.test.ts

@@ -140,20 +140,20 @@ describe('context-pruner handleSteps', () => {
     const firstTerminalMessage = resultMessages.find(
       (m: any) =>
         m.role === 'tool' &&
-        m.content?.toolName === 'run_terminal_command' &&
-        m.content?.output?.[0]?.value?.command === 'command-1',
+        m.toolName === 'run_terminal_command' &&
+        m.content?.[0]?.value?.command === 'command-1',
     )
     expect(
-      firstTerminalMessage?.content?.output?.[0]?.value?.stdoutOmittedForLength,
+      firstTerminalMessage?.content?.[0]?.value?.stdoutOmittedForLength,
     ).toBe(true)
 
     // Check that recent terminal commands are preserved (but may be processed by large tool result pass)
     const recentTerminalMessage = resultMessages.find(
       (m: any) =>
         m.role === 'tool' &&
-        m.content?.toolName === 'run_terminal_command' &&
-        (m.content?.output?.[0]?.value?.command === 'command-7' ||
-          m.content?.output?.[0]?.value?.message ===
+        m.toolName === 'run_terminal_command' &&
+        (m.content?.[0]?.value?.command === 'command-7' ||
+          m.content?.[0]?.value?.message ===
             '[LARGE_TOOL_RESULT_OMITTED]'),
     )
     expect(recentTerminalMessage).toBeDefined()
@@ -181,17 +181,17 @@ describe('context-pruner handleSteps', () => {
 
     // Large tool result should be simplified
     const largeResultMessage = resultMessages.find(
-      (m: any) => m.role === 'tool' && m.content?.toolName === 'read_files',
+      (m: any) => m.role === 'tool' && m.toolName === 'read_files',
     )
-    expect(largeResultMessage?.content?.output?.[0]?.value?.message).toBe(
+    expect(largeResultMessage?.content?.[0]?.value?.message).toBe(
       '[LARGE_TOOL_RESULT_OMITTED]',
     )
 
     // Small tool result should be preserved
     const smallResultMessage = resultMessages.find(
-      (m: any) => m.role === 'tool' && m.content?.toolName === 'code_search',
+      (m: any) => m.role === 'tool' && m.toolName === 'code_search',
     )
-    expect(smallResultMessage?.content?.output?.[0]?.value?.data).toBe(
+    expect(smallResultMessage?.content?.[0]?.value?.data).toBe(
       'Small result',
     )
   })
@@ -367,7 +367,7 @@ describe('context-pruner edge cases', () => {
     // Valid terminal command should be processed correctly
     const validCommand = resultMessages.find(
       (m: any) =>
-        m.role === 'tool' && m.content?.toolName === 'run_terminal_command',
+        m.role === 'tool' && m.toolName === 'run_terminal_command',
     )
     expect(validCommand).toBeDefined()
   })
@@ -481,7 +481,7 @@ describe('context-pruner edge cases', () => {
     const hasLargeToolResultReplacement = resultMessages.some(
       (m: any) =>
         m.role === 'tool' &&
-        m.content?.output?.[0]?.value?.message ===
+        m.content?.[0]?.value?.message ===
           '[LARGE_TOOL_RESULT_OMITTED]',
     )
     expect(hasLargeToolResultReplacement).toBe(true)

.agents/base2/base2.ts

@@ -404,7 +404,7 @@ function buildImplementationStepPrompt({
     !isFast &&
       `You must spawn the ${useGeminiEditor ? 'editor-implementor-gemini' : isMax ? 'editor-best-of-n-gpt-5' : 'editor-best-of-n'} agent to implement code changes, since it will generate the best code changes.`,
     isMax && 'Spawn the thinker-best-of-n-gpt-5 to solve complex problems.',
-    `After completing the user request, summarize your changes in a sentence${isFast ? '' : ' or a few short bullet points'}.${isSonnet ? " Don't create any summary markdown files or example documentation files, unless asked by the user." : ''}. Don't repeat yourself -- especially if you already summarized your changes then just end your turn.`,
+    `After completing the user request, summarize your changes in a sentence${isFast ? '' : ' or a few short bullet points'}.${isSonnet ? " Don't create any summary markdown files or example documentation files, unless asked by the user." : ''}.`,
   ).join('\n')
 }
 

.agents/file-explorer/file-picker.ts

@@ -1,4 +1,4 @@
-import { ToolCall } from 'types/agent-definition'
+import { StepText, ToolCall } from 'types/agent-definition'
 import { publisher } from '../constants'
 
 import {
@@ -50,7 +50,7 @@ In your report, please give a very concise analysis that includes the full paths
 Do not use any further tools or spawn any further agents.
   `.trim(),
 
-  handleSteps: function* ({ prompt, params }) {
+  handleSteps: function* ({ prompt, params, logger }) {
     const { toolResult: fileListerResults } = yield {
       toolName: 'spawn_agents',
       input: {
@@ -64,21 +64,46 @@ Do not use any further tools or spawn any further agents.
       },
     } satisfies ToolCall
 
-    const fileListerResult = fileListerResults?.[0]
-    const filesStr =
-      fileListerResult && fileListerResult.type === 'json'
-        ? ((fileListerResult.value as any)?.[0]?.value?.value as string)
-        : ''
-    const files = filesStr.split('\n').filter(Boolean)
+    const filesResult =
+      extractSpawnResults<{ text: string }[]>(fileListerResults)[0]
+    if (!Array.isArray(filesResult)) {
+      yield {
+        type: 'STEP_TEXT',
+        text: filesResult.errorMessage,
+      } satisfies StepText
+      return
+    }
+
+    const paths = filesResult[0].text.split('\n').filter(Boolean)
 
     yield {
       toolName: 'read_files',
       input: {
-        paths: files,
+        paths,
       },
     }
 
     yield 'STEP'
+
+    function extractSpawnResults<T>(
+      results: any[] | undefined,
+    ): (T | { errorMessage: string })[] {
+      if (!results) return []
+      const spawnedResults = results
+        .filter((result) => result.type === 'json')
+        .map((result) => result.value)
+        .flat() as {
+        agentType: string
+        value: { value?: T; errorMessage?: string }
+      }[]
+      return spawnedResults.map(
+        (result) =>
+          result.value.value ?? {
+            errorMessage:
+              result.value.errorMessage ?? 'Error extracting spawn results',
+          },
+      )
+    }
   },
 }
 

cli/package.json

@@ -15,17 +15,13 @@
     }
   },
   "scripts": {
-    "dev": "./scripts/dev.sh",
-    "prebuild": "bun run build:sdk",
-    "build": "bun build src/index.tsx --outdir dist --target bun --format esm",
+    "dev": "bun run src/index.tsx",
     "build:sdk": "cd ../sdk && bun run build",
-    "build:sdk-types": "cd ../sdk && bun run build:types",
     "build:binary": "bun ./scripts/build-binary.ts codebuff $npm_package_version",
     "release": "bun run scripts/release.ts",
     "start": "bun run dist/index.js",
     "test": "bun test",
     "test:tmux-poc": "bun run src/__tests__/tmux-poc.ts",
-    "pretypecheck": "bun run build:sdk-types",
     "typecheck": "tsc --noEmit -p ."
   },
   "sideEffects": false,

cli/release/package.json

@@ -1,18 +1,19 @@
 {
   "name": "codebuff",
-  "version": "1.0.522",
+  "version": "1.0.523",
   "description": "AI coding agent",
   "license": "MIT",
   "bin": {
     "codebuff": "index.js",
     "cb": "index.js"
   },
   "scripts": {
-    "postinstall": "node -e \"const fs = require('fs'); const path = require('path'); const os = require('os'); const binaryPath = path.join(os.homedir(), '.config', 'manicode', process.platform === 'win32' ? 'codebuff.exe' : 'codebuff'); try { fs.unlinkSync(binaryPath) } catch (e) { /* ignore if file doesn't exist */ }\"",
+    "postinstall": "node postinstall.js",
     "preuninstall": "node -e \"const fs = require('fs'); const path = require('path'); const os = require('os'); const binaryPath = path.join(os.homedir(), '.config', 'manicode', process.platform === 'win32' ? 'codebuff.exe' : 'codebuff'); try { fs.unlinkSync(binaryPath) } catch (e) { /* ignore if file doesn't exist */ }\""
   },
   "files": [
     "index.js",
+    "postinstall.js",
     "README.md"
   ],
   "os": [

cli/release/postinstall.js

@@ -0,0 +1,34 @@
+#!/usr/bin/env node
+
+const fs = require('fs');
+const path = require('path');
+const os = require('os');
+
+// Clean up old binary
+const binaryPath = path.join(
+  os.homedir(),
+  '.config',
+  'manicode',
+  process.platform === 'win32' ? 'codebuff.exe' : 'codebuff'
+);
+
+try {
+  fs.unlinkSync(binaryPath);
+} catch (e) {
+  /* ignore if file doesn't exist */
+}
+
+// Print welcome message
+console.log('\n');
+console.log('🎉 Welcome to Codebuff!');
+console.log('\n');
+console.log('To get started:');
+console.log('  1. cd to your project directory');
+console.log('  2. Run: codebuff');
+console.log('\n');
+console.log('Example:');
+console.log('  $ cd ~/my-project');
+console.log('  $ codebuff');
+console.log('\n');
+console.log('For more information, visit: https://codebuff.com/docs');
+console.log('\n');

cli/scripts/dev.sh

@@ -151,6 +151,23 @@ describe('API Integration', () => {
       // 401s are now logged as auth failures
       expect(testLogger.error.mock.calls.length).toBeGreaterThan(0)
     })
+
+    test('should treat 404 from /api/v1/me as invalid credentials (AuthenticationError)', async () => {
+      setFetchMock(async () => {
+        return new Response(null, { status: 404 })
+      })
+      const testLogger = createLoggerMocks()
+
+      await expect(
+        getUserInfoFromApiKey({
+          apiKey: 'not-found-token',
+          fields: ['id'],
+          logger: testLogger,
+        }),
+      ).rejects.toBeInstanceOf(AuthenticationError)
+
+      expect(testLogger.error.mock.calls.length).toBeGreaterThan(0)
+    })
   })
 
   describe('P1: Error Response Handling', () => {