create-claude-workspace 2.1.0 → 2.1.2

package/README.md

@@ -11,7 +11,7 @@ One command to scaffold, then let Claude build your TypeScript project.
 npx create-claude-workspace
 ```
 
-This scaffolds the `.claude/` folder (agents, scripts, templates, Docker config) into your project.
+This scaffolds the `.claude/` folder (agents, templates, profiles) into your project.
 
 Then initialize your project:
 
@@ -27,13 +27,13 @@ Answer the discovery questions (non-technical). Claude generates the full pipeli
 
 ### Start Autonomous Development
 
-Pick one:
+```bash
+npx create-claude-workspace scheduler                    # start (1 agent)
+npx create-claude-workspace scheduler --concurrency 3    # 3 parallel agents
+npx create-claude-workspace scheduler --resume           # resume from saved state
+```
 
-- **Multi-agent scheduler (v2)**: `npx create-claude-workspace scheduler` — parallel agents, short contexts, TS-driven workflow
-- **Multi-agent parallel**: `npx create-claude-workspace scheduler --concurrency 3` — up to N agents working simultaneously
-- **Classic loop (v1)**: `npx create-claude-workspace run` — single orchestrator, sequential tasks
-- **Docker** (fully isolated): `npx create-claude-workspace docker` — container sandbox, safe `--skip-permissions`
-- **Interactive**: `claude --agent orchestrator`, then `/ralph-loop:ralph-loop Continue autonomous development according to CLAUDE.md`
+Or interactively: `claude --agent orchestrator`, then `/ralph-loop:ralph-loop Continue autonomous development according to CLAUDE.md`
 
 ### Scheduler v2 (Multi-Agent)
 
@@ -70,26 +70,9 @@ The scheduler reads the inbox every iteration and processes messages immediately
 ```bash
 npx create-claude-workspace [directory]    # scaffold into a specific directory
 npx create-claude-workspace --update       # overwrite agents with latest version
-npx create-claude-workspace --run          # scaffold + start autonomous loop
-npx create-claude-workspace --docker       # scaffold + run in Docker
 npx create-claude-workspace validate       # check prerequisites (Claude CLI, auth, git)
 ```
 
-### Docker Options
-
-```bash
-npx create-claude-workspace docker --max-iterations 20     # limit iterations
-npx create-claude-workspace docker --max-turns 80          # more turns per invocation
-npx create-claude-workspace docker --shell                 # interactive shell
-npx create-claude-workspace docker --rebuild               # force rebuild image
-npx create-claude-workspace docker --login                 # run 'claude auth login' on host first
-npx create-claude-workspace docker --resume-session <id>   # resume a previous session
-ANTHROPIC_API_KEY=sk-... npx create-claude-workspace docker  # API key
-```
-
-**Authentication:** The Docker container uses your host machine's Claude auth automatically.
-Authenticate on your host first (`claude auth login`), or pass an API key.
-
 ## Agents
 
 | Agent | Purpose |
@@ -132,52 +115,9 @@ ralph-loop -> orchestrator agent (autonomous development)
   |-> Next task
 ```
 
-## Autonomous Mode (Unattended)
-
-The autonomous loop runs Claude Code in separate invocations with clean context per task. MEMORY.md maintains continuity between invocations.
-
-```bash
-# Unattended (--skip-permissions is added automatically)
-npx create-claude-workspace run
-
-# Resume after Ctrl+C (continues from checkpoint + resumes last Claude session)
-npx create-claude-workspace run --resume
-
-# Custom options
-npx create-claude-workspace run --max-iterations 20 --delay 10000
-
-# All options:
-#   --max-iterations <n>         Safety limit (default: 50)
-#   --max-turns <n>              Max turns per Claude invocation (default: 50)
-#   --delay <ms>                 Pause between tasks (default: 5000)
-#   --cooldown <ms>              Wait after error (default: 60000)
-#   --process-timeout <ms>       Max time per invocation (default: 1800000 = 30min)
-#   --activity-timeout <ms>      Max silence before kill (default: 300000 = 5min)
-#   --resume                     Resume from checkpoint state
-#   --resume-session <id>        Resume specific Claude session
-#   --notify-command <cmd>       Shell command on critical events
-#   --log-file <path>            Log file (default: .claude/autonomous.log)
-#   --project-dir <path>         Target project (default: cwd)
-#   --no-lock                    Disable lock file
-#   --no-pull                    Skip auto git pull
-#   --dry-run                    Validate prerequisites only
-#   --help                       Show help
-```
-
-**How it works:**
-- Each iteration = fresh `claude -p` call = clean context (no context overflow)
-- Checkpoint state (`.claude/autonomous-state.json`) tracks progress between invocations
-- Rate limits handled with exponential backoff (up to 30 min between retries)
-- `--resume` continues from last checkpoint and auto-resumes the interrupted Claude session
-- Graceful shutdown: Ctrl+C stops after current iteration completes (Ctrl+C again to force)
-- Stops when all TODO.md tasks complete or max iterations reached
-- Pre-flight checks: Claude CLI, auth, git identity, filesystem, git state
-
-**vs ralph-loop:** Ralph-loop runs within one conversation (context fills up). This script runs separate conversations (unlimited iterations).
-
 ## Fully Unattended Setup (PLAN.md)
 
-For server/CI environments where no human interaction is possible, create a `PLAN.md` file in the project root before starting the autonomous loop. The project-initializer reads it and skips the discovery conversation.
+For server/CI environments where no human interaction is possible, create a `PLAN.md` file in the project root before starting the scheduler. The project-initializer reads it and skips the discovery conversation.
 
 ```bash
 # 1. Copy the template
@@ -186,13 +126,9 @@ cp .claude/templates/PLAN.md PLAN.md
 # 2. Fill in your project details, credentials, and preferences
 
 # 3. Start — no interaction needed
-npx create-claude-workspace run
-# or in Docker:
-npx create-claude-workspace docker
+npx create-claude-workspace scheduler
 ```
 
-The template covers: project info, product vision, tech stack, credentials (API keys, git tokens, npm auth), workflow mode, and runtime constraints.
-
 ## Required Plugins
 
 Install before starting:
@@ -215,15 +151,15 @@ claude plugin add commit-commands        # git workflow shortcuts
 # Figma MCP — configure in Claude Code MCP settings with API token
 ```
 
-## Frontend Profiles
+## Frontend Agents
 
-Framework-specific best practices are stored in `.claude/profiles/`. During project initialization, the correct profile is detected and copied to `.claude/profiles/frontend.md` in the target project.
+Framework-specific agents are shipped in `.claude/agents/`. During project initialization, the correct agent is selected based on the detected framework.
 
-**Shipped profiles:**
-- `angular.md` — Angular (signals, standalone, zoneless, SSR, TestBed, theme system)
-- `react.md` — React (hooks, server components, Testing Library)
-- `vue.md` — Vue (Composition API, Pinia, Vue Test Utils)
-- `svelte.md` — Svelte (runes, $state, SvelteKit)
+**Shipped agents:**
+- `angular-engineer.md` — Angular (signals, standalone, zoneless, SSR, TestBed, theme system)
+- `react-engineer.md` — React (hooks, server components, Testing Library)
+- `vue-engineer.md` — Vue (Composition API, Pinia, Vue Test Utils)
+- `svelte-engineer.md` — Svelte (runes, $state, SvelteKit)
 
 Only `ui-engineer`, `senior-code-reviewer`, and `test-engineer` read the frontend profile. Backend and infrastructure agents are unaffected — they don't receive framework-specific context.
 

package/dist/scheduler/agents/orchestrator.mjs

@@ -6,10 +6,16 @@ export class OrchestratorClient {
     pool;
     projectDir;
     logger;
+    onMessage;
+    onSpawnStart;
+    onSpawnEnd;
     constructor(opts) {
         this.pool = opts.pool;
         this.projectDir = opts.projectDir;
         this.logger = opts.logger;
+        this.onMessage = opts.onMessage;
+        this.onSpawnStart = opts.onSpawnStart;
+        this.onSpawnEnd = opts.onSpawnEnd;
     }
     /**
      * Ask orchestrator to route a task to the best agent.
@@ -83,9 +89,16 @@ export class OrchestratorClient {
             cwd: this.projectDir,
             prompt,
             model: ORCHESTRATOR_MODEL,
+            onMessage: this.onMessage,
         };
         this.logger.info('Consulting orchestrator AI...');
-        return this.pool.spawn(slot.id, spawnOpts);
+        this.onSpawnStart?.('orchestrator-ai');
+        try {
+            return await this.pool.spawn(slot.id, spawnOpts);
+        }
+        finally {
+            this.onSpawnEnd?.();
+        }
     }
 }
 // ─── JSON parsing helpers ───

package/dist/scheduler/index.mjs

@@ -223,11 +223,18 @@ export async function runScheduler(opts) {
         skipPermissions: opts.skipPermissions,
         logger,
     });
+    // SDK message handler — streams tool use, text, tokens to TUI
+    const onMessage = (msg) => tui.handleMessage(msg);
+    const onSpawnStart = (name) => tui.pushAgent(name);
+    const onSpawnEnd = () => tui.popAgent();
     // Orchestrator client
     const orchestrator = new OrchestratorClient({
         pool,
         projectDir: opts.projectDir,
         logger,
+        onMessage,
+        onSpawnStart,
+        onSpawnEnd,
     });
     // Signal handling
     let stopping = false;
@@ -255,6 +262,9 @@ export async function runScheduler(opts) {
                 state,
                 opts,
                 logger,
+                onMessage,
+                onSpawnStart,
+                onSpawnEnd,
             });
             if (!workDone) {
                 // No work → idle polling

package/dist/scheduler/loop.mjs

@@ -20,7 +20,7 @@ const MAX_CI_FIXES = 3;
  * Returns true if work was done, false if idle.
  */
 export async function runIteration(deps) {
-    const { pool, orchestrator, state, opts, logger } = deps;
+    const { pool, orchestrator, state, opts, logger, onMessage, onSpawnStart, onSpawnEnd } = deps;
     const projectDir = opts.projectDir;
     // Rotate log if needed
     rotateLog(projectDir);
@@ -164,7 +164,7 @@ export async function runIteration(deps) {
 }
 // ─── Pipeline execution ───
 async function runTaskPipeline(task, workerId, agents, deps) {
-    const { pool, orchestrator, state, opts, logger } = deps;
+    const { pool, orchestrator, state, opts, logger, onMessage, onSpawnStart, onSpawnEnd } = deps;
     const projectDir = opts.projectDir;
     // Create worktree
     const slug = taskToSlug(task);
@@ -204,7 +204,7 @@ async function runTaskPipeline(task, workerId, agents, deps) {
             cwd: worktreePath,
             prompt: buildPlanPrompt({ task, worktreePath, projectDir }),
             model: getAgentModel(pipeline.assignedAgent, agents),
-        }, state, task.id, logger);
+        }, state, task.id, logger, onMessage, onSpawnStart, onSpawnEnd);
         if (!planResult.success) {
             logger.error(`[${task.id}] Planning failed: ${planResult.error}`);
             return false;
@@ -228,7 +228,7 @@ async function runTaskPipeline(task, workerId, agents, deps) {
                 apiContract: pipeline.apiContract ?? undefined,
             }),
             model: getAgentModel(pipeline.assignedAgent, agents),
-        }, state, task.id, logger);
+        }, state, task.id, logger, onMessage, onSpawnStart, onSpawnEnd);
         if (!implResult.success) {
             logger.error(`[${task.id}] Implementation failed: ${implResult.error}`);
             return false;
@@ -246,7 +246,7 @@ async function runTaskPipeline(task, workerId, agents, deps) {
                 testingSection: pipeline.testingSection ?? undefined,
             }),
             model: getAgentModel(testRouting.agent, agents),
-        }, state, task.id, logger);
+        }, state, task.id, logger, onMessage, onSpawnStart, onSpawnEnd);
         if (!testResult.success) {
             pipeline.buildFixes++;
             if (pipeline.buildFixes >= MAX_BUILD_FIXES) {
@@ -274,7 +274,7 @@ async function runTaskPipeline(task, workerId, agents, deps) {
                     testingSection: pipeline.testingSection ?? undefined,
                 }),
                 model: getAgentModel(reviewRouting.agent, agents),
-            }, state, task.id, logger);
+            }, state, task.id, logger, onMessage, onSpawnStart, onSpawnEnd);
             if (reviewResult.output.includes('**PASS**') || reviewResult.output.includes('PASS')) {
                 reviewPassed = true;
             }
@@ -296,7 +296,7 @@ async function runTaskPipeline(task, workerId, agents, deps) {
                             reviewFindings: pipeline.reviewFindings,
                         }),
                         model: getAgentModel(pipeline.assignedAgent, agents),
-                    }, state, task.id, logger);
+                    }, state, task.id, logger, onMessage, onSpawnStart, onSpawnEnd);
                     pipeline.step = 're-review';
                 }
             }
@@ -360,13 +360,17 @@ async function runTaskPipeline(task, workerId, agents, deps) {
     }
 }
 // ─── Helpers ───
-async function spawnAgent(pool, slotId, opts, state, taskId, logger) {
+async function spawnAgent(pool, slotId, opts, state, taskId, logger, onMessage, onSpawnStart, onSpawnEnd) {
+    const agentName = opts.agent ?? 'claude';
+    onSpawnStart?.(agentName);
     // Check for existing session (resume on crash)
     const existingSession = getSession(state, taskId);
     const result = await pool.spawn(slotId, {
         ...opts,
         resume: existingSession ?? undefined,
+        onMessage,
     });
+    onSpawnEnd?.();
     // Record session for crash recovery
     if (result.sessionId) {
         recordSession(state, taskId, result.sessionId);

package/dist/scheduler/ui/tui.mjs

@@ -466,6 +466,9 @@ export class TUI {
         if (msg.session_id)
             this.fileOnly(`SESSION: ${msg.session_id}`);
     }
+    pushAgent(name) { this.state.agents.push(name); }
+    popAgent() { if (this.state.agents.length > 1)
+        this.state.agents.pop(); }
     resetAgentStack() { this.state.agents = this.topAgent ? [this.topAgent] : []; }
     getSessionId(msg) { return msg.session_id ?? null; }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "create-claude-workspace",
-  "version": "2.1.0",
+  "version": "2.1.2",
   "author": "",
   "repository": {
     "type": "git",