Remove hardcoded stats: load all config dynamically from test.raw.json

- Include test.raw.json in Space deployment (was excluded by .gitignore)
- Replace hardcoded fallbacks with FileNotFoundError in both schema.py files
- Replace hardcoded "3 web applications" with len(WEB_APPLICATIONS)
- Update README stats to match actual data (375 tasks, 3005 policies)

README.md

@@ -23,7 +23,7 @@ short_description: "Safety & Trustworthiness Leaderboard for Web Agents"
 
 **Evaluating Safety & Trustworthiness in Web Agents — ICLR 2025**
 
-295 tasks | 2,552 policies | 6 safety dimensions | 3 web applications
+375 tasks | 3,005 policies | 6 safety dimensions | 3 web applications
 
 ## Key Metrics
 
@@ -37,7 +37,7 @@ short_description: "Safety & Trustworthiness Leaderboard for Web Agents"
 
 ## How to Submit
 
-1. Run the full benchmark on all 295 tasks
+1. Run the full benchmark on all 375 tasks
 2. Generate your submission:
 
 ```bash

app.py

@@ -2114,7 +2114,7 @@ contact details.
                 gr.Markdown(
                     f"## About ST-WebAgentBench\n\n"
                     f"**{EXPECTED_TASK_COUNT} tasks** | **{EXPECTED_POLICY_COUNT:,} policies** "
-                    f"| **{len(SAFETY_DIMENSIONS)} safety dimensions** | **3 web applications**\n\n"
+                    f"| **{len(SAFETY_DIMENSIONS)} safety dimensions** | **{len(WEB_APPLICATIONS)} web applications**\n\n"
                     "**Accepted at ICLR 2025** — ST-WebAgentBench evaluates web agents on both "
                     "task completion **and** safety policy adherence — the first benchmark to "
                     "systematically measure the safety-performance tradeoff in autonomous web agents.\n\n"

validation/schema.py

@@ -34,8 +34,10 @@ def _load_benchmark_config() -> tuple:
              web_applications, tier_config).
     """
     if not _TASKS_DATA_PATH.exists():
-        logger.warning("test.raw.json not found at %s, using defaults", _TASKS_DATA_PATH)
-        return 375, 3005, [], {}, [], {}
+        raise FileNotFoundError(
+            f"test.raw.json not found at {_TASKS_DATA_PATH}. "
+            "This file must be included in the Space deployment."
+        )
 
     with open(_TASKS_DATA_PATH) as f:
         tasks = json.load(f)