diff --git a/.claude/settings.json b/.claude/settings.json
new file mode 100644
index 0000000000..6d308befa7
--- /dev/null
+++ b/.claude/settings.json
@@ -0,0 +1,10 @@
+{
+  "permissions": {
+    "allowedTools": [
+      "Read", "Grep", "Glob",
+      "Bash(ls:*)", "Bash(cat:*)", "Bash(grep:*)", "Bash(find:*)",
+      "Bash(git status:*)", "Bash(git diff:*)", "Bash(git log:*)", "Bash(git worktree:*)",
+      "Bash(tmux:*)", "Bash(sleep:*)", "Bash(branchlet:*)"
+    ]
+  }
+}
diff --git a/.github/workflows/classic-forge-ci.yml b/.github/workflows/classic-forge-ci.yml
index 07ceec4f90..fa0aa41981 100644
--- a/.github/workflows/classic-forge-ci.yml
+++ b/.github/workflows/classic-forge-ci.yml
@@ -55,6 +55,9 @@ jobs:
       - name: Install Python dependencies
         run: poetry install
 
+      - name: Install Playwright browsers
+        run: poetry run playwright install chromium
+
       - name: Run pytest with coverage
         run: |
           poetry run pytest -vv \
diff --git a/classic/benchmark/.env.example b/classic/benchmark/.env.example
new file mode 100644
index 0000000000..e4fb03486f
--- /dev/null
+++ b/classic/benchmark/.env.example
@@ -0,0 +1,4 @@
+AGENT_NAME=mini-agi
+REPORTS_FOLDER="reports/mini-agi"
+OPENAI_API_KEY="sk-" # for LLM eval
+BUILD_SKILL_TREE=false # set to true to build the skill tree.
diff --git a/classic/benchmark/frontend/.env.example b/classic/benchmark/frontend/.env.example
new file mode 100644
index 0000000000..168cf5b05c
--- /dev/null
+++ b/classic/benchmark/frontend/.env.example
@@ -0,0 +1,14 @@
+# Since the ".env" file is gitignored, you can use the ".env.example" file to
+# build a new ".env" file when you clone the repo. Keep this file up-to-date
+# when you add new variables to `.env`.
+
+# This file will be committed to version control, so make sure not to have any
+# secrets in it. If you are cloning this repo, create a copy of this file named
+# ".env" and populate it with your secrets.
+
+# When adding additional environment variables, the schema in "/src/env.mjs"
+# should be updated accordingly.
+
+# Prisma
+# https://www.prisma.io/docs/reference/database-reference/connection-urls#env
+DATABASE_URL="file:./db.sqlite"
diff --git a/classic/direct_benchmark/direct_benchmark/__main__.py b/classic/direct_benchmark/direct_benchmark/__main__.py
index f855ba08b9..d5d9731303 100644
--- a/classic/direct_benchmark/direct_benchmark/__main__.py
+++ b/classic/direct_benchmark/direct_benchmark/__main__.py
@@ -19,6 +19,22 @@ from .models import (
 from .ui import console
 
 
+def get_default_model() -> str:
+    """Get the default model based on available API keys.
+
+    Returns the model preset name for the first available API key,
+    preferring Claude > OpenAI > Groq.
+    """
+    if os.environ.get("ANTHROPIC_API_KEY"):
+        return "claude"
+    elif os.environ.get("OPENAI_API_KEY"):
+        return "openai"
+    elif os.environ.get("GROQ_API_KEY"):
+        return "groq"
+    # Fallback to openai (most commonly available in CI)
+    return "openai"
+
+
 @click.group()
 @click.version_option(version="0.1.0")
 def cli():
@@ -40,8 +56,8 @@ def cli():
 @click.option(
     "--models",
     "-m",
-    default="claude",
-    help=f"Comma-separated model presets. Available: {', '.join(MODEL_PRESETS.keys())}",
+    default=None,
+    help=f"Comma-separated model presets. Auto-detects from API keys if not specified. Available: {', '.join(MODEL_PRESETS.keys())}",
 )
 @click.option(
     "--categories",
@@ -232,7 +248,7 @@ def cli():
 )
 def run(
     strategies: str,
-    models: str,
+    models: Optional[str],
     categories: Optional[str],
     skip_categories: Optional[str],
     tests: Optional[str],
@@ -280,7 +296,11 @@ def run(
         console.print(f"Available: {STRATEGIES}")
         sys.exit(1)
 
-    # Parse models
+    # Parse models (auto-detect from API keys if not specified)
+    if models is None:
+        models = get_default_model()
+        console.print(f"[dim]Auto-detected model: {models}[/dim]")
+
     model_list = [m.strip() for m in models.split(",")]
     invalid_models = [m for m in model_list if m not in MODEL_PRESETS]
     if invalid_models: