Merge pull request #226 from AutoForgeAI/feat/batch-size-limits-and-testing-batch-setting

feat: increase batch size limits to 15 and add testing_batch_size setting

Author: leonvanzyl

CLAUDE.md

@@ -65,7 +65,7 @@ python autonomous_agent_demo.py --project-dir my-app --yolo
 # Parallel mode: run multiple agents concurrently (1-5 agents)
 python autonomous_agent_demo.py --project-dir my-app --parallel --max-concurrency 3
 
-# Batch mode: implement multiple features per agent session (1-3)
+# Batch mode: implement multiple features per agent session (1-15)
 python autonomous_agent_demo.py --project-dir my-app --batch-size 3
 
 # Batch specific features by ID
@@ -496,9 +496,9 @@ The orchestrator enforces strict bounds on concurrent processes:
 
 ### Multi-Feature Batching
 
-Agents can implement multiple features per session using `--batch-size` (1-3, default: 3):
+Agents can implement multiple features per session using `--batch-size` (1-15, default: 3):
 - `--batch-size N` - Max features per coding agent batch
-- `--testing-batch-size N` - Features per testing batch (1-5, default: 3)
+- `--testing-batch-size N` - Features per testing batch (1-15, default: 3)
 - `--batch-features 1,2,3` - Specific feature IDs for batch implementation
 - `--testing-batch-features 1,2,3` - Specific feature IDs for batch regression testing
 - `prompts.py` provides `get_batch_feature_prompt()` for multi-feature prompt generation

autonomous_agent_demo.py

@@ -176,14 +176,14 @@ def parse_args() -> argparse.Namespace:
         "--testing-batch-size",
         type=int,
         default=3,
-        help="Number of features per testing batch (1-5, default: 3)",
+        help="Number of features per testing batch (1-15, default: 3)",
     )
 
     parser.add_argument(
         "--batch-size",
         type=int,
         default=3,
-        help="Max features per coding agent batch (1-3, default: 3)",
+        help="Max features per coding agent batch (1-15, default: 3)",
     )
 
     return parser.parse_args()

package.json

@@ -1,6 +1,6 @@
 {
   "name": "autoforge-ai",
-  "version": "0.1.16",
+  "version": "0.1.17",
   "description": "Autonomous coding agent with web UI - build complete apps with AI",
   "license": "AGPL-3.0",
   "bin": {

parallel_orchestrator.py

@@ -131,7 +131,7 @@ def _dump_database_state(feature_dicts: list[dict], label: str = ""):
 MAX_PARALLEL_AGENTS = 5
 MAX_TOTAL_AGENTS = 10
 DEFAULT_CONCURRENCY = 3
-DEFAULT_TESTING_BATCH_SIZE = 3  # Number of features per testing batch (1-5)
+DEFAULT_TESTING_BATCH_SIZE = 3  # Number of features per testing batch (1-15)
 POLL_INTERVAL = 5  # seconds between checking for ready features
 MAX_FEATURE_RETRIES = 3  # Maximum times to retry a failed feature
 INITIALIZER_TIMEOUT = 1800  # 30 minutes timeout for initializer
@@ -168,7 +168,7 @@ def __init__(
             yolo_mode: Whether to run in YOLO mode (skip testing agents entirely)
             testing_agent_ratio: Number of regression testing agents to maintain (0-3).
                 0 = disabled, 1-3 = maintain that many testing agents running independently.
-            testing_batch_size: Number of features to include per testing session (1-5).
+            testing_batch_size: Number of features to include per testing session (1-15).
                 Each testing agent receives this many features to regression test.
             on_output: Callback for agent output (feature_id, line)
             on_status: Callback for agent status changes (feature_id, status)
@@ -178,8 +178,8 @@ def __init__(
         self.model = model
         self.yolo_mode = yolo_mode
         self.testing_agent_ratio = min(max(testing_agent_ratio, 0), 3)  # Clamp 0-3
-        self.testing_batch_size = min(max(testing_batch_size, 1), 5)  # Clamp 1-5
-        self.batch_size = min(max(batch_size, 1), 3)  # Clamp 1-3
+        self.testing_batch_size = min(max(testing_batch_size, 1), 15)  # Clamp 1-15
+        self.batch_size = min(max(batch_size, 1), 15)  # Clamp 1-15
         self.on_output = on_output
         self.on_status = on_status
 

server/routers/agent.py

@@ -17,11 +17,11 @@
 from ..utils.validation import validate_project_name
 
 
-def _get_settings_defaults() -> tuple[bool, str, int, bool, int]:
+def _get_settings_defaults() -> tuple[bool, str, int, bool, int, int]:
     """Get defaults from global settings.
 
     Returns:
-        Tuple of (yolo_mode, model, testing_agent_ratio, playwright_headless, batch_size)
+        Tuple of (yolo_mode, model, testing_agent_ratio, playwright_headless, batch_size, testing_batch_size)
     """
     import sys
     root = Path(__file__).parent.parent.parent
@@ -47,7 +47,12 @@ def _get_settings_defaults() -> tuple[bool, str, int, bool, int]:
     except (ValueError, TypeError):
         batch_size = 3
 
-    return yolo_mode, model, testing_agent_ratio, playwright_headless, batch_size
+    try:
+        testing_batch_size = int(settings.get("testing_batch_size", "3"))
+    except (ValueError, TypeError):
+        testing_batch_size = 3
+
+    return yolo_mode, model, testing_agent_ratio, playwright_headless, batch_size, testing_batch_size
 
 
 router = APIRouter(prefix="/api/projects/{project_name}/agent", tags=["agent"])
@@ -96,14 +101,15 @@ async def start_agent(
     manager = get_project_manager(project_name)
 
     # Get defaults from global settings if not provided in request
-    default_yolo, default_model, default_testing_ratio, playwright_headless, default_batch_size = _get_settings_defaults()
+    default_yolo, default_model, default_testing_ratio, playwright_headless, default_batch_size, default_testing_batch_size = _get_settings_defaults()
 
     yolo_mode = request.yolo_mode if request.yolo_mode is not None else default_yolo
     model = request.model if request.model else default_model
     max_concurrency = request.max_concurrency or 1
     testing_agent_ratio = request.testing_agent_ratio if request.testing_agent_ratio is not None else default_testing_ratio
 
     batch_size = default_batch_size
+    testing_batch_size = default_testing_batch_size
 
     success, message = await manager.start(
         yolo_mode=yolo_mode,
@@ -112,6 +118,7 @@ async def start_agent(
         testing_agent_ratio=testing_agent_ratio,
         playwright_headless=playwright_headless,
         batch_size=batch_size,
+        testing_batch_size=testing_batch_size,
     )
 
     # Notify scheduler of manual start (to prevent auto-stop during scheduled window)

server/routers/settings.py

@@ -113,6 +113,7 @@ async def get_settings():
         testing_agent_ratio=_parse_int(all_settings.get("testing_agent_ratio"), 1),
         playwright_headless=_parse_bool(all_settings.get("playwright_headless"), default=True),
         batch_size=_parse_int(all_settings.get("batch_size"), 3),
+        testing_batch_size=_parse_int(all_settings.get("testing_batch_size"), 3),
         api_provider=api_provider,
         api_base_url=all_settings.get("api_base_url"),
         api_has_auth_token=bool(all_settings.get("api_auth_token")),
@@ -138,6 +139,9 @@ async def update_settings(update: SettingsUpdate):
     if update.batch_size is not None:
         set_setting("batch_size", str(update.batch_size))
 
+    if update.testing_batch_size is not None:
+        set_setting("testing_batch_size", str(update.testing_batch_size))
+
     # API provider settings
     if update.api_provider is not None:
         old_provider = get_setting("api_provider", "claude")
@@ -177,6 +181,7 @@ async def update_settings(update: SettingsUpdate):
         testing_agent_ratio=_parse_int(all_settings.get("testing_agent_ratio"), 1),
         playwright_headless=_parse_bool(all_settings.get("playwright_headless"), default=True),
         batch_size=_parse_int(all_settings.get("batch_size"), 3),
+        testing_batch_size=_parse_int(all_settings.get("testing_batch_size"), 3),
         api_provider=api_provider,
         api_base_url=all_settings.get("api_base_url"),
         api_has_auth_token=bool(all_settings.get("api_auth_token")),

server/schemas.py

@@ -444,7 +444,8 @@ class SettingsResponse(BaseModel):
     ollama_mode: bool = False  # True when api_provider is "ollama"
     testing_agent_ratio: int = 1  # Regression testing agents (0-3)
     playwright_headless: bool = True
-    batch_size: int = 3  # Features per coding agent batch (1-3)
+    batch_size: int = 3  # Features per coding agent batch (1-15)
+    testing_batch_size: int = 3  # Features per testing agent batch (1-15)
     api_provider: str = "claude"
     api_base_url: str | None = None
     api_has_auth_token: bool = False  # Never expose actual token
@@ -463,7 +464,8 @@ class SettingsUpdate(BaseModel):
     model: str | None = None
     testing_agent_ratio: int | None = None  # 0-3
     playwright_headless: bool | None = None
-    batch_size: int | None = None  # Features per agent batch (1-3)
+    batch_size: int | None = None  # Features per agent batch (1-15)
+    testing_batch_size: int | None = None  # Features per testing agent batch (1-15)
     api_provider: str | None = None
     api_base_url: str | None = Field(None, max_length=500)
     api_auth_token: str | None = Field(None, max_length=500)  # Write-only, never returned
@@ -500,8 +502,15 @@ def validate_testing_ratio(cls, v: int | None) -> int | None:
     @field_validator('batch_size')
     @classmethod
     def validate_batch_size(cls, v: int | None) -> int | None:
-        if v is not None and (v < 1 or v > 3):
-            raise ValueError("batch_size must be between 1 and 3")
+        if v is not None and (v < 1 or v > 15):
+            raise ValueError("batch_size must be between 1 and 15")
+        return v
+
+    @field_validator('testing_batch_size')
+    @classmethod
+    def validate_testing_batch_size(cls, v: int | None) -> int | None:
+        if v is not None and (v < 1 or v > 15):
+            raise ValueError("testing_batch_size must be between 1 and 15")
         return v
 
 

server/services/process_manager.py

@@ -374,6 +374,7 @@ async def start(
         testing_agent_ratio: int = 1,
         playwright_headless: bool = True,
         batch_size: int = 3,
+        testing_batch_size: int = 3,
     ) -> tuple[bool, str]:
         """
         Start the agent as a subprocess.
@@ -440,6 +441,9 @@ async def start(
         # Add --batch-size flag for multi-feature batching
         cmd.extend(["--batch-size", str(batch_size)])
 
+        # Add --testing-batch-size flag for testing agent batching
+        cmd.extend(["--testing-batch-size", str(testing_batch_size)])
+
         # Apply headless setting to .playwright/cli.config.json so playwright-cli
         # picks it up (the only mechanism it supports for headless control)
         self._apply_playwright_headless(playwright_headless)

ui/package-lock.json

@@ -10,6 +10,7 @@ import {
   DialogTitle,
 } from '@/components/ui/dialog'
 import { Switch } from '@/components/ui/switch'
+import { Slider } from '@/components/ui/slider'
 import { Label } from '@/components/ui/label'
 import { Alert, AlertDescription } from '@/components/ui/alert'
 import { Button } from '@/components/ui/button'
@@ -63,6 +64,12 @@ export function SettingsModal({ isOpen, onClose }: SettingsModalProps) {
     }
   }
 
+  const handleTestingBatchSizeChange = (size: number) => {
+    if (!updateSettings.isPending) {
+      updateSettings.mutate({ testing_batch_size: size })
+    }
+  }
+
   const handleProviderChange = (providerId: string) => {
     if (!updateSettings.isPending) {
       updateSettings.mutate({ api_provider: providerId })
@@ -432,28 +439,34 @@ export function SettingsModal({ isOpen, onClose }: SettingsModalProps) {
               </div>
             </div>
 
-            {/* Features per Agent */}
+            {/* Features per Coding Agent */}
             <div className="space-y-2">
-              <Label className="font-medium">Features per Agent</Label>
+              <Label className="font-medium">Features per Coding Agent</Label>
               <p className="text-sm text-muted-foreground">
-                Number of features assigned to each coding agent
+                Number of features assigned to each coding agent session
               </p>
-              <div className="flex rounded-lg border overflow-hidden">
-                {[1, 2, 3].map((size) => (
-                  <button
-                    key={size}
-                    onClick={() => handleBatchSizeChange(size)}
-                    disabled={isSaving}
-                    className={`flex-1 py-2 px-3 text-sm font-medium transition-colors ${
-                      (settings.batch_size ?? 1) === size
-                        ? 'bg-primary text-primary-foreground'
-                        : 'bg-background text-foreground hover:bg-muted'
-                    } ${isSaving ? 'opacity-50 cursor-not-allowed' : ''}`}
-                  >
-                    {size}
-                  </button>
-                ))}
-              </div>
+              <Slider
+                min={1}
+                max={15}
+                value={settings.batch_size ?? 3}
+                onChange={handleBatchSizeChange}
+                disabled={isSaving}
+              />
+            </div>
+
+            {/* Features per Testing Agent */}
+            <div className="space-y-2">
+              <Label className="font-medium">Features per Testing Agent</Label>
+              <p className="text-sm text-muted-foreground">
+                Number of features assigned to each testing agent session
+              </p>
+              <Slider
+                min={1}
+                max={15}
+                value={settings.testing_batch_size ?? 3}
+                onChange={handleTestingBatchSizeChange}
+                disabled={isSaving}
+              />
             </div>
 
             {/* Update Error */}