feat: add pre-submission compliance checker

check_compliance.py

@@ -0,0 +1,111 @@
+"""
+CodeSensei Pre-Submission Compliance Checker
+Validates inference.py against ALL hackathon requirements
+"""
+import os
+import sys
+
+REPO = r"x:\scalerXmeta"
+INF = os.path.join(REPO, "inference.py")
+
+results = []
+
+def check(name, condition):
+    status = "PASS" if condition else "FAIL"
+    results.append((name, status))
+    print(f"  {'✅' if condition else '❌'} {name}: {status}")
+    return condition
+
+print("=" * 60)
+print("CODESENSEI SUBMISSION COMPLIANCE CHECK")
+print("=" * 60)
+
+# --- File existence ---
+print("\n📁 FILE STRUCTURE:")
+check("inference.py exists in root", os.path.isfile(INF))
+check("openenv.yaml exists", os.path.isfile(os.path.join(REPO, "openenv.yaml")))
+check("Dockerfile exists in root", os.path.isfile(os.path.join(REPO, "Dockerfile")))
+check("requirements.txt exists", os.path.isfile(os.path.join(REPO, "requirements.txt")))
+check("env/data/bug_dataset.json exists", os.path.isfile(os.path.join(REPO, "env", "data", "bug_dataset.json")))
+
+# --- Read inference.py ---
+with open(INF, "r", encoding="utf-8") as f:
+    code = f.read()
+    lines = code.split("\n")
+
+# --- OpenAI Client ---
+print("\n🔌 OPENAI CLIENT:")
+check("from openai import OpenAI", "from openai import OpenAI" in code)
+check("client.chat.completions.create used", "client.chat.completions.create" in code)
+
+# --- Env Vars ---
+print("\n🔑 ENV VARS:")
+check("API_BASE_URL has default", 'os.getenv("API_BASE_URL"' in code)
+check("MODEL_NAME has default", 'os.getenv("MODEL_NAME"' in code)
+check("HF_TOKEN has NO default", 'os.getenv("HF_TOKEN")' in code)
+
+# Check HF_TOKEN does NOT have a default value
+hf_lines = [l for l in lines if "HF_TOKEN" in l and "getenv" in l]
+hf_no_default = any("HF_TOKEN\")" in l or "HF_TOKEN')" in l for l in hf_lines)
+check("HF_TOKEN truly has no default (no 2nd arg)", hf_no_default)
+
+check("LOCAL_IMAGE_NAME / IMAGE_NAME", "IMAGE_NAME" in code)
+
+# --- Stdout Logging ---
+print("\n📋 STDOUT LOGGING FORMAT:")
+check("[START] log line exists", "[START]" in code)
+check("[STEP] log line exists", "[STEP]" in code)
+check("[END] log line exists", "[END]" in code)
+
+# Check exact format
+check("[START] has task= env= model=", "task={" in code and "env={" in code and "model={" in code)
+check("[STEP] has step= action= reward= done= error=", 
+      "step={" in code and "action={" in code and "reward={" in code and "done={" in code and "error={" in code)
+check("[END] has success= steps= score= rewards=",
+      "success={" in code and "steps={" in code and "score={" in code and "rewards={" in code)
+
+# Check formatting
+check("reward formatted to 2dp ({reward:.2f})", "reward:.2f" in code)
+check("score formatted to 3dp ({score:.3f})", "score:.3f" in code)
+check("done lowercase (str(done).lower())", "str(done).lower()" in code)
+check("success lowercase (str(success).lower())", "str(success).lower()" in code)
+
+# --- OpenEnv Pattern ---
+print("\n🐳 OPENENV PATTERN:")
+check("from_docker_image() used", "from_docker_image" in code)
+check("await env.reset()", "await env.reset()" in code)
+check("await env.step(", "await env.step(" in code)
+check("result.observation accessed", "result.observation" in code)
+check("result.done accessed", "result.done" in code)
+check("result.reward accessed", "result.reward" in code)
+check("await env.close() in finally", "await env.close()" in code)
+check("asyncio.run(main())", "asyncio.run(main())" in code)
+
+# --- Score ---
+print("\n📊 SCORE NORMALIZATION:")
+check("Score clamped to [0, 1]", "min(max(score" in code or "max(0.0, min(1.0" in code)
+
+# --- requirements.txt ---
+print("\n📦 DEPENDENCIES:")
+with open(os.path.join(REPO, "requirements.txt"), "r") as f:
+    reqs = f.read()
+check("openai in requirements.txt", "openai" in reqs)
+
+# --- Fallback return ---
+print("\n🔄 EDGE CASES:")
+check('Fallback returns "hello"', '"hello"' in code)
+check("Exception handled in model call", "except Exception" in code)
+
+# --- Summary ---
+print("\n" + "=" * 60)
+passed = sum(1 for _, s in results if s == "PASS")
+total = len(results)
+if passed == total:
+    print(f"🎉 ALL {total}/{total} CHECKS PASSED!")
+    print("   Your submission is READY!")
+else:
+    failed = [(n, s) for n, s in results if s == "FAIL"]
+    print(f"⚠️  {passed}/{total} PASSED, {len(failed)} FAILED:")
+    for name, _ in failed:
+        print(f"   ❌ {name}")
+print("=" * 60)