Make Lite into Free mode!

Author: jahooma

agents/base2/base2-free.ts

@@ -0,0 +1,8 @@
+import { createBase2 } from './base2'
+
+const definition = {
+  ...createBase2('free'),
+  id: 'base2-free',
+  displayName: 'Buffy the Free Orchestrator',
+}
+export default definition

agents/base2/base2-lite.ts

@@ -7,7 +7,7 @@ import {
 } from '../types/secret-agent-definition'
 
 export function createBase2(
-  mode: 'default' | 'lite' | 'max' | 'fast',
+  mode: 'default' | 'free' | 'max' | 'fast',
   options?: {
     hasNoValidation?: boolean
     planOnly?: boolean
@@ -22,15 +22,15 @@ export function createBase2(
   const isDefault = mode === 'default'
   const isFast = mode === 'fast'
   const isMax = mode === 'max'
-  const isLite = mode === 'lite'
+  const isFree = mode === 'free'
 
-  const isOpus = !isLite
+  const isOpus = !isFree
   const isSonnet = false
   const isGemini = false
 
   return {
     publisher,
-    model: isLite ? 'x-ai/grok-4.1-fast' : 'anthropic/claude-opus-4.5',
+    model: isFree ? 'x-ai/grok-4.1-fast' : 'anthropic/claude-opus-4.5',
     displayName: 'Buffy the Orchestrator',
     spawnerPrompt:
       'Advanced base agent that orchestrates planning, editing, and reviewing for complex coding tasks',
@@ -55,7 +55,7 @@ export function createBase2(
       'spawn_agents',
       'read_files',
       'read_subtree',
-      !isFast && !isLite && 'write_todos',
+      !isFast && !isFree && 'write_todos',
       !isFast && !noAskUser && 'suggest_followups',
       'str_replace',
       'write_file',
@@ -72,11 +72,11 @@ export function createBase2(
       'glob-matcher',
       'researcher-web',
       'researcher-docs',
-      isLite ? 'commander-lite' : 'commander',
+      isFree ? 'commander-lite' : 'commander',
       isDefault && 'thinker',
       (isDefault || isMax) && ['opus-agent', 'gpt-5-agent'],
       isMax && 'thinker-best-of-n-opus',
-      isLite && 'editor-glm',
+      isFree && 'editor-glm',
       isDefault && 'editor',
       isMax && 'editor-multi-prompt',
       isDefault && 'code-reviewer',
@@ -133,7 +133,7 @@ Use the spawn_agents tool to spawn specialized agents to help you complete the u
 - **Sequence agents properly:** Keep in mind dependencies when spawning different agents. Don't spawn agents in parallel that depend on each other.
   ${buildArray(
         '- Spawn context-gathering agents (file pickers, code-searcher, directory-lister, glob-matcher, and web/docs researchers) before making edits.',
-        isLite &&
+        isFree &&
         '- Spawn the editor-glm agent to implement the changes after you have gathered all the context you need.',
         isDefault &&
         '- Spawn the editor agent to implement the changes after you have gathered all the context you need.',
@@ -198,7 +198,7 @@ ${isDefault
         ? `[ You implement the changes using the editor agent ]`
         : isFast
           ? '[ You implement the changes using the str_replace or write_file tools ]'
-          : isLite
+          : isFree
             ? '[ You implement the changes using the editor-glm agent ]'
             : '[ You implement the changes using the editor-multi-prompt agent ]'
       }
@@ -248,7 +248,7 @@ ${PLACEHOLDER.GIT_CHANGES_PROMPT}
         isFast,
         isDefault,
         isMax,
-        isLite,
+        isFree,
         hasNoValidation,
         noAskUser,
       }),
@@ -260,7 +260,7 @@ ${PLACEHOLDER.GIT_CHANGES_PROMPT}
         isMax,
         hasNoValidation,
         isSonnet,
-        isLite,
+        isFree,
         noAskUser,
       }),
 
@@ -292,15 +292,15 @@ function buildImplementationInstructionsPrompt({
   isFast,
   isDefault,
   isMax,
-  isLite,
+  isFree,
   hasNoValidation,
   noAskUser,
 }: {
   isSonnet: boolean
   isFast: boolean
   isDefault: boolean
   isMax: boolean
-  isLite: boolean
+  isFree: boolean
   hasNoValidation: boolean
   noAskUser: boolean
 }) {
@@ -320,7 +320,7 @@ ${buildArray(
     `- For any task requiring 3+ steps, use the write_todos tool to write out your step-by-step implementation plan. Include ALL of the applicable tasks in the list.${isFast ? '' : ' You should include a step to review the changes after you have implemented the changes.'}:${hasNoValidation ? '' : ' You should include at least one step to validate/test your changes: be specific about whether to typecheck, run tests, run lints, etc.'} You may be able to do reviewing and validation in parallel in the same step. Skip write_todos for simple tasks like quick edits or answering questions.`,
     (isDefault || isMax) &&
     `- For quick problems, briefly explain your reasoning to the user. If you need to think longer, write your thoughts within the <think> tags. Finally, for complex problems, spawn the thinker agent to help find the best solution. (gpt-5-agent is a last resort for complex problems)`,
-    isLite &&
+    isFree &&
     '- IMPORTANT: You must spawn the editor-glm agent to implement the changes after you have gathered all the context you need. This agent will do the best job of implementing the changes so you must spawn it for all changes. Do not pass any prompt or params to the editor agent when spawning it. It will make its own best choices of what to do.',
     isDefault &&
     '- IMPORTANT: You must spawn the editor agent to implement the changes after you have gathered all the context you need. This agent will do the best job of implementing the changes so you must spawn it for all non-trivial changes. Do not pass any prompt or params to the editor agent when spawning it. It will make its own best choices of what to do.',
@@ -347,15 +347,15 @@ function buildImplementationStepPrompt({
   isMax,
   hasNoValidation,
   isSonnet,
-  isLite,
+  isFree,
   noAskUser,
 }: {
   isDefault: boolean
   isFast: boolean
   isMax: boolean
   hasNoValidation: boolean
   isSonnet: boolean
-  isLite: boolean
+  isFree: boolean
   noAskUser: boolean
 }) {
   return buildArray(

agents/base2/base2.ts

@@ -17,7 +17,7 @@ describe('AgentModeToggle - buildExpandedSegments', () => {
   for (const mode of modes) {
     test(`returns segments with active indicator for ${mode}`, () => {
       const segs = buildExpandedSegments(mode)
-      // 4 mode options (DEFAULT, LITE, MAX, PLAN) + 1 active indicator
+      // 4 mode options (DEFAULT, FREE, MAX, PLAN) + 1 active indicator
       expect(segs.length).toBe(5)
 
       // Current mode is disabled among the choices

cli/src/__tests__/unit/agent-mode-toggle.test.ts

@@ -42,12 +42,10 @@ export const ThinkingBlock = memo(
       }
     }, [onToggleCollapsed, thinkingId])
 
-    // thinkingOpen === true means still streaming
-    // thinkingOpen === false means explicitly closed with </think> tag
-    // thinkingOpen === undefined means native reasoning block - complete when message is complete
+    // thinkingOpen === false means explicitly closed (with </think> tag or message completion)
+    // Otherwise (true or undefined), completion is determined by message completion
     const isThinkingComplete =
-      firstBlock?.thinkingOpen === false ||
-      (firstBlock?.thinkingOpen === undefined && isMessageComplete)
+      firstBlock?.thinkingOpen === false || isMessageComplete
 
     // Hide if no content or no thinkingId (but NOT when thinking is complete)
     if (!combinedContent || !thinkingId) {

cli/src/components/blocks/thinking-block.tsx

@@ -4,7 +4,7 @@ import { setCurrentChatId } from '../project-files'
 import { createStreamController } from './stream-state'
 import { useChatStore } from '../state/chat-store'
 import { getCodebuffClient } from '../utils/codebuff-client'
-import { AGENT_MODE_TO_ID } from '../utils/constants'
+import { AGENT_MODE_TO_ID, AGENT_MODE_TO_COST_MODE } from '../utils/constants'
 import { createEventHandlerState } from '../utils/create-event-handler-state'
 import { createRunConfig } from '../utils/create-run-config'
 import { loadAgentDefinitions } from '../utils/local-agent-registry'
@@ -443,6 +443,7 @@ export const useSendMessage = ({
           agentDefinitions,
           eventHandlerState,
           signal: abortController.signal,
+          costMode: AGENT_MODE_TO_COST_MODE[agentMode],
         })
 
         logger.info({ runConfig }, '[send-message] Sending message with sdk run config')

cli/src/hooks/use-send-message.ts

@@ -114,7 +114,8 @@ function parseArgs(): ParsedArgs {
       '--cwd <directory>',
       'Set the working directory (default: current directory)',
     )
-    .option('--lite', 'Start in LITE mode')
+    .option('--free', 'Start in FREE mode')
+    .option('--lite', 'Start in FREE mode (deprecated, use --free)')
     .option('--max', 'Start in MAX mode')
     .option('--plan', 'Start in PLAN mode')
     .helpOption('-h, --help', 'Show this help message')
@@ -129,7 +130,7 @@ function parseArgs(): ParsedArgs {
 
   // Determine initial mode from flags (last flag wins if multiple specified)
   let initialMode: AgentMode | undefined
-  if (options.lite) initialMode = 'LITE'
+  if (options.free || options.lite) initialMode = 'FREE'
   if (options.max) initialMode = 'MAX'
   if (options.plan) initialMode = 'PLAN'
 
@@ -148,6 +149,7 @@ function parseArgs(): ParsedArgs {
 }
 
 async function main(): Promise<void> {
+  console.log()
   // Run OSC theme detection BEFORE anything else.
   // This MUST happen before OpenTUI starts because OSC responses come through stdin,
   // and OpenTUI also listens to stdin. Running detection here ensures stdin is clean.

cli/src/index.tsx

@@ -107,10 +107,21 @@ export const MAIN_AGENT_ID = 'main-agent'
  */
 export const AGENT_MODE_TO_ID = {
   DEFAULT: 'base2',
-  LITE: 'base2-lite',
+  FREE: 'base2-free',
   MAX: 'base2-max',
   PLAN: 'base2-plan',
 } as const
 
 export type AgentMode = keyof typeof AGENT_MODE_TO_ID
 export const AGENT_MODES = Object.keys(AGENT_MODE_TO_ID) as AgentMode[]
+
+/**
+ * Maps CLI agent mode to cost mode for billing.
+ * FREE mode maps to 'free' cost mode where all agents cost 0 credits.
+ */
+export const AGENT_MODE_TO_COST_MODE = {
+  DEFAULT: 'normal',
+  FREE: 'free',
+  MAX: 'max',
+  PLAN: 'normal',
+} as const satisfies Record<AgentMode, 'free' | 'normal' | 'max' | 'experimental' | 'ask'>

cli/src/utils/constants.ts

@@ -23,6 +23,7 @@ export type CreateRunConfigParams = {
   agentDefinitions: AgentDefinition[]
   eventHandlerState: EventHandlerState
   signal: AbortSignal
+  costMode?: 'free' | 'normal' | 'max' | 'experimental' | 'ask'
 }
 
 const SENSITIVE_EXTENSIONS = new Set([
@@ -98,6 +99,7 @@ export const createRunConfig = (params: CreateRunConfigParams) => {
     previousRunState,
     agentDefinitions,
     eventHandlerState,
+    costMode,
   } = params
 
   return {
@@ -111,6 +113,7 @@ export const createRunConfig = (params: CreateRunConfigParams) => {
     handleStreamChunk: createStreamChunkHandler(eventHandlerState),
     handleEvent: createEventHandler(eventHandlerState),
     signal: params.signal,
+    costMode,
     fileFilter: ((filePath: string) => {
       if (isSensitiveFile(filePath)) return { status: 'blocked' }
       if (isEnvTemplateFile(filePath)) return { status: 'allow-example' }

cli/src/utils/create-run-config.ts

@@ -1,4 +1,4 @@
-import type { ChatMessage, ContentBlock } from '../types/chat'
+import type { ChatMessage, ContentBlock, TextContentBlock } from '../types/chat'
 
 // Small wrapper to avoid repeating the ai-message map/update pattern.
 export type SetMessagesFn = (
@@ -57,9 +57,25 @@ export const createMessageUpdater = (
   const markComplete = (metadata?: Partial<ChatMessage>) => {
     updateAiMessage((msg) => {
       const { metadata: messageMetadata, ...rest } = metadata ?? {}
+      
+      // Mark native reasoning blocks as complete by setting thinkingOpen = false
+      // This ensures thinking blocks auto-collapse when the message finishes
+      // Check for thinkingOpen !== false to handle both true (native) and undefined (legacy)
+      const updatedBlocks = msg.blocks?.map((block) => {
+        if (
+          block.type === 'text' &&
+          (block as TextContentBlock).textType === 'reasoning' &&
+          (block as TextContentBlock).thinkingOpen !== false
+        ) {
+          return { ...block, thinkingOpen: false } as ContentBlock
+        }
+        return block
+      })
+      
       const nextMessage: ChatMessage = {
         ...msg,
         isComplete: true,
+        ...(updatedBlocks && { blocks: updatedBlocks }),
         ...rest,
       }
 
@@ -184,9 +200,25 @@ export const createBatchedMessageUpdater = (
       prev.map((msg) => {
         if (msg.id !== aiMessageId) return msg
         const { metadata: messageMetadata, ...rest } = metadata ?? {}
+        
+        // Mark native reasoning blocks as complete by setting thinkingOpen = false
+        // This ensures thinking blocks auto-collapse when the message finishes
+        // Check for thinkingOpen !== false to handle both true (native) and undefined (legacy)
+        const updatedBlocks = msg.blocks?.map((block) => {
+          if (
+            block.type === 'text' &&
+            (block as TextContentBlock).textType === 'reasoning' &&
+            (block as TextContentBlock).thinkingOpen !== false
+          ) {
+            return { ...block, thinkingOpen: false } as ContentBlock
+          }
+          return block
+        })
+        
         const nextMessage: ChatMessage = {
           ...msg,
           isComplete: true,
+          ...(updatedBlocks && { blocks: updatedBlocks }),
           ...rest,
         }
         if (messageMetadata) {