fix: improve AI feature install error handling and resource limits

Handle OOM kills (exit code 137) with actionable memory guidance,
filter ANSI/progress noise from error output, add --no-cache-dir to
pip installs, reduce download concurrency to 2, and bump default
container memory from 4g to 6g.

apps/api/src/routes/features.ts

@@ -225,13 +225,24 @@ export async function registerFeatureRoutes(app: FastifyInstance): Promise<void>
             }
           }
           if (!errorMsg) {
-            const meaningful = lastStderrLines.filter(
+            if (code === 137) {
-              (l) => !l.startsWith("{") && !l.includes("pthread_setaffinity_np"),
+              errorMsg =
-            );
+                "Installation was killed due to insufficient memory. " +
-            errorMsg =
+                "Try increasing the container's memory limit (e.g. mem_limit: 6g in docker-compose.yml) and retry.";
-              meaningful.join("\n") ||
+            } else {
-              stdoutBuffer.trim() ||
+              const meaningful = lastStderrLines.filter(
-              `Install failed with exit code ${code}`;
+                (l) =>
+                  !l.startsWith("{") &&
+                  !l.includes("pthread_setaffinity_np") &&
+                  !l.includes("\x1b[") &&
+                  !l.includes("━") &&
+                  !/^\s*\d+%\|/.test(l),
+              );
+              errorMsg =
+                meaningful.join("\n") ||
+                stdoutBuffer.trim() ||
+                `Install failed with exit code ${code}`;
+            }
           }
           setInstallProgress(bundleId, null, errorMsg);
           updateSingleFileProgress({ jobId, phase: "failed", percent: 0, error: errorMsg });

docker/docker-compose.yml

@@ -52,8 +52,8 @@ services:
       # - COOKIE_SECRET=
     restart: unless-stopped
     # --- Security hardening ---
-    mem_limit: 4g
+    mem_limit: 6g
-    memswap_limit: 4g
+    memswap_limit: 6g
     cpus: 4
     pids_limit: 512
     cap_drop:

packages/ai/python/install_feature.py

@@ -125,7 +125,7 @@ def check_disk_space(path: str, min_bytes: int = 100 * 1024 * 1024) -> None:
 
 def pip_install(package: str, extra_flags: list[str] | None = None) -> None:
     """Run pip install for a single package spec. Raises on failure."""
-    cmd = [sys.executable, "-m", "pip", "install"]
+    cmd = [sys.executable, "-m", "pip", "install", "--no-cache-dir"]
     if extra_flags:
         cmd.extend(extra_flags)
 
@@ -471,7 +471,7 @@ def download_models(models: list[dict], models_dir: str) -> list[str]:
         except Exception as e:
             return (model_id, e)
 
-    with concurrent.futures.ThreadPoolExecutor(max_workers=4) as pool:
+    with concurrent.futures.ThreadPoolExecutor(max_workers=2) as pool:
         futures = {
             pool.submit(_download, i, m): i
             for i, m in enumerate(models)