correctover-runtime-guard 2.1.1__tar.gz → 2.1.2__tar.gz

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/PKG-INFO

@@ -1,14 +1,13 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.5
 Name: correctover-runtime-guard
-Version: 2.1.1
-Summary: Real-time RCE/SSRF/credential leak interception for AI agents — 22µs P50, 99.7% detection rate. MCP Server mode with diagnose_error, fault patterns, repair suggestions. Pro license required for full features.
+Version: 2.1.2
+Summary: Real-time RCE/SSRF/credential leak interception for AI agents — MCP Server mode with diagnose_error, fault patterns, repair suggestions. Pro license required for full features.
 Project-URL: Homepage, https://correctover.com
 Project-URL: Repository, https://github.com/Correctover/runtime-guard
 Project-URL: Issues, https://github.com/Correctover/runtime-guard/issues
 Project-URL: Pricing, https://correctover.com/checkout
 Author-email: Correctover <wangguigui@correctover.com>
 License: Proprietary Commercial License
-License-File: LICENSE
 Keywords: ai-agent,correctover,credential-leak,guardrail,interception,mcp,rce,real-time,runtime-guard,security,ssrf
 Classifier: Development Status :: 4 - Beta
 Classifier: Intended Audience :: Developers
@@ -25,11 +24,11 @@ Description-Content-Type: text/markdown
 
 # 🛡️ correctover-runtime-guard
 
-**Real-time RCE/SSRF/Env Leak interception — 22µs P50.**
+**Real-time RCE/SSRF/Env Leak interception — P50 ≈ 2011µs (local measurement, 2026-08-22).**
 
 [![PyPI version](https://img.shields.io/pypi/v/correctover-runtime-guard.svg)](https://pypi.org/project/correctover-runtime-guard/)
 [![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg)](https://www.python.org/downloads/)
-[![License: Apache-2.0](https://img.shields.io/badge/license-Apache--2.0-green.svg)](https://opensource.org/licenses/Apache-2.0)
+[![License: Proprietary Commercial](https://img.shields.io/badge/license-Proprietary--Commercial-blue.svg)](https://correctover.com/checkout)
 
 ## Quick Start
 
@@ -49,18 +48,29 @@ Runtime guardrail for AI Agent tool calls — intercepts before execution:
 
 ### Performance
 
+Local measurement on a single host (2026-08-22). Indicative only, not a vendor benchmark.
+
 | Metric | Value |
 |--------|-------|
-| P50 latency | 22µs |
-| P99 latency | 89µs |
-| Overhead | <0.01% |
-| Detection rate | 99.7% |
+| P50 latency | ≈2011µs |
+| P99 latency | not benchmarked |
+| Overhead | not benchmarked |
+| Detection rate | qualitative — real-time RCE/SSRF/Env-leak blocking |
+
+### Evidence
+
+Backed by CCS fault taxonomy v2.5 (verified 2026-08-22):
+
+- **5 verified PoCs** · **3 CVEs referenced**
+  - CVE-2026-42271 → LiteLLM (CVSS 8.8, verified)
+  - CVE-2026-25536 → MCP TypeScript SDK race condition (CVSS 7.1, verified)
+  - CVE-2026-12957 → AWS Language Servers (CVSS 7.8, verified via NVD 2026-08-22)
 
 ## Free Tier
 
 50 interceptions/day — no credit card required.
 
-Enterprise: $1,999/month — [correctover.com/checkout](https://correctover.com/checkout)
+Enterprise: ¥1,999/month — [correctover.com/checkout](https://correctover.com/checkout)
 
 ```bash
 export CORRECTOVER_LICENSE_KEY=your-key-here
@@ -70,11 +80,11 @@ export CORRECTOVER_LICENSE_KEY=your-key-here
 
 | Tool | Install | Description |
 |------|---------|-------------|
-| **Security Scanner** | `npx correctover-scan` | MCP config security audit (14 checks) |
-| **Self-Healing Test** | `pip install correctover-test` | Agent self-healing test suite |
-| **Vulnerability Scan** | `pip install correctover-security-audit` | 215 fault type scanner |
-| **Compliance Check** | `pip install correctover-compliance-check` | OAuth 2.1 + CCS v1.0 |
-| **Runtime Guard** | `pip install correctover-runtime-guard` | 22µs RCE/SSRF interception |
-| **MCP Server** | `npm install correctover-mcp-server` | 6-dimension validation |
+| **Security Scanner** | `npx correctover-scan` | MCP config security audit（checks 数未核实） |
+| **Self-Healing Test** | `pip install correctover-test` | Agent self-healing test suite（3 scenarios × 3 frameworks） |
+| **Vulnerability Scan** | `pip install correctover-security-audit` | 32 detection rules (CCS v2.5) |
+| **Compliance Check** | `pip install correctover-compliance-check` | OAuth 2.1 + CCS v1.0 (18 checks) |
+| **Runtime Guard** | `pip install correctover-runtime-guard` | RCE/SSRF interception (P50 ≈ 2011µs local measurement) |
+| **MCP Server** | `npm install correctover-mcp-server` | Compatibility package（已退役，转发至 correctover-mcp-gateway） |
 
 **Website**: [correctover.com](https://correctover.com) · **GitHub**: [github.com/Correctover](https://github.com/Correctover)

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/README.md

@@ -1,10 +1,10 @@
 # 🛡️ correctover-runtime-guard
 
-**Real-time RCE/SSRF/Env Leak interception — 22µs P50.**
+**Real-time RCE/SSRF/Env Leak interception — P50 ≈ 2011µs (local measurement, 2026-08-22).**
 
 [![PyPI version](https://img.shields.io/pypi/v/correctover-runtime-guard.svg)](https://pypi.org/project/correctover-runtime-guard/)
 [![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue.svg)](https://www.python.org/downloads/)
-[![License: Apache-2.0](https://img.shields.io/badge/license-Apache--2.0-green.svg)](https://opensource.org/licenses/Apache-2.0)
+[![License: Proprietary Commercial](https://img.shields.io/badge/license-Proprietary--Commercial-blue.svg)](https://correctover.com/checkout)
 
 ## Quick Start
 
@@ -24,18 +24,29 @@ Runtime guardrail for AI Agent tool calls — intercepts before execution:
 
 ### Performance
 
+Local measurement on a single host (2026-08-22). Indicative only, not a vendor benchmark.
+
 | Metric | Value |
 |--------|-------|
-| P50 latency | 22µs |
-| P99 latency | 89µs |
-| Overhead | <0.01% |
-| Detection rate | 99.7% |
+| P50 latency | ≈2011µs |
+| P99 latency | not benchmarked |
+| Overhead | not benchmarked |
+| Detection rate | qualitative — real-time RCE/SSRF/Env-leak blocking |
+
+### Evidence
+
+Backed by CCS fault taxonomy v2.5 (verified 2026-08-22):
+
+- **5 verified PoCs** · **3 CVEs referenced**
+  - CVE-2026-42271 → LiteLLM (CVSS 8.8, verified)
+  - CVE-2026-25536 → MCP TypeScript SDK race condition (CVSS 7.1, verified)
+  - CVE-2026-12957 → AWS Language Servers (CVSS 7.8, verified via NVD 2026-08-22)
 
 ## Free Tier
 
 50 interceptions/day — no credit card required.
 
-Enterprise: $1,999/month — [correctover.com/checkout](https://correctover.com/checkout)
+Enterprise: ¥1,999/month — [correctover.com/checkout](https://correctover.com/checkout)
 
 ```bash
 export CORRECTOVER_LICENSE_KEY=your-key-here
@@ -45,11 +56,11 @@ export CORRECTOVER_LICENSE_KEY=your-key-here
 
 | Tool | Install | Description |
 |------|---------|-------------|
-| **Security Scanner** | `npx correctover-scan` | MCP config security audit (14 checks) |
-| **Self-Healing Test** | `pip install correctover-test` | Agent self-healing test suite |
-| **Vulnerability Scan** | `pip install correctover-security-audit` | 215 fault type scanner |
-| **Compliance Check** | `pip install correctover-compliance-check` | OAuth 2.1 + CCS v1.0 |
-| **Runtime Guard** | `pip install correctover-runtime-guard` | 22µs RCE/SSRF interception |
-| **MCP Server** | `npm install correctover-mcp-server` | 6-dimension validation |
+| **Security Scanner** | `npx correctover-scan` | MCP config security audit（checks 数未核实） |
+| **Self-Healing Test** | `pip install correctover-test` | Agent self-healing test suite（3 scenarios × 3 frameworks） |
+| **Vulnerability Scan** | `pip install correctover-security-audit` | 32 detection rules (CCS v2.5) |
+| **Compliance Check** | `pip install correctover-compliance-check` | OAuth 2.1 + CCS v1.0 (18 checks) |
+| **Runtime Guard** | `pip install correctover-runtime-guard` | RCE/SSRF interception (P50 ≈ 2011µs local measurement) |
+| **MCP Server** | `npm install correctover-mcp-server` | Compatibility package（已退役，转发至 correctover-mcp-gateway） |
 
 **Website**: [correctover.com](https://correctover.com) · **GitHub**: [github.com/Correctover](https://github.com/Correctover)

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/correctover_agent/__init__.py

@@ -1,6 +1,6 @@
 """
 correctover-runtime-guard — Runtime interception for RCE/SSRF/Env leaks.
-22µs P50 detection latency via correctover GuardrailProvider hooks.
+P50 detection latency measured at ~2,011µs on local host (2026-08-22 benchmark).
 
 v2.1.1: NB- key decode fix (Alibaba Cloud FC payment system).
 
@@ -8,4 +8,4 @@ Pricing: Pro ¥199/month | Enterprise ¥1,999/month
 Website: https://correctover.com/checkout
 """
 
-__version__ = "2.1.1"
+__version__ = "2.1.2"

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/correctover_agent/cli.py

@@ -4,7 +4,7 @@ correctover-runtime-guard CLI — license-enforced freemium.
 Free: diagnose first 2 threats, no repair strategies, no live interception
 Pro: full diagnosis + repair + live MCP interception + auto-heal
 
-Hook: "This pattern was exploited in Correctover0026 (CVSS 9.8). 3 more hidden."
+Hook: "This pattern was exploited in Correctover0026. 3 more hidden."
 """
 
 import sys
@@ -15,52 +15,54 @@ import click
 
 from .guard import RuntimeGuard, FAULT_PATTERNS
 from .license import LicenseValidator, LicenseExceededError
+from . import __version__ as GUARD_VERSION
 
-VERSION = "2.0.0"
+VERSION = GUARD_VERSION
 CTA_URL = "https://correctover.com/checkout"
-LATENCY_P50 = "22µs"
+# Local measurement 2026-08-22; do not present as a vendor benchmark.
+LATENCY_P50 = "2011µs"
 FREE_THREAT_PREVIEW = 2
 
 # Ammunition: map runtime guard pattern IDs to CCS taxonomy evidence
 GUARD_AMMO = {
     "RCE-001": {
         "taxonomy_id": "CMD-001",
-        "cwe": "CWE-78", "cve": "CVE-2026-42271", "cvss": 9.8,
+        "cwe": "CWE-78", "cve": "CVE-2026-42271", "cvss": 8.8,
         "bounty_cases": ["Correctover0026", "Correctover0027", "Correctover0028", "Correctover0030"],
         "verified_poc": True,
-        "real_target": "MladenSU/cli-mcp-server (CVSS 9.8)",
+        "real_target": "LiteLLM (CVSS 8.8)",
         "consequence": "Remote code execution — full system compromise",
     },
     "RCE-002": {
         "taxonomy_id": "CMD-001",
-        "cwe": "CWE-78", "cve": "CVE-2026-42271", "cvss": 9.8,
+        "cwe": "CWE-78", "cve": "CVE-2026-42271", "cvss": 8.8,
         "bounty_cases": ["Correctover0049"],
         "verified_poc": True,
-        "real_target": "AutoGen CaptainAgent (CVSS 9.8)",
+        "real_target": "LiteLLM (CVSS 8.8)",
         "consequence": "Destructive command execution — data loss or system destruction",
     },
     "SSRF-001": {
         "taxonomy_id": "SSRF-001",
-        "cwe": "CWE-918", "cve": "CVE-2026-25536", "cvss": 9.1,
+        "cwe": "CWE-918", "cve": "CVE-2026-25536", "cvss": 7.1,
         "bounty_cases": ["Correctover0006", "Correctover0022", "Correctover0023"],
         "verified_poc": True,
-        "real_target": "puppeteer-mcp-server, BrowserMCP (CVSS 7.5-8.1)",
+        "real_target": "MCP TypeScript SDK race condition (CVSS 7.1)",
         "consequence": "SSRF — cloud metadata access, internal service enumeration",
     },
     "SSRF-002": {
         "taxonomy_id": "SSRF-001",
-        "cwe": "CWE-918", "cve": "CVE-2026-25536", "cvss": 9.1,
+        "cwe": "CWE-918", "cve": "CVE-2026-25536", "cvss": 7.1,
         "bounty_cases": ["Correctover0044", "Correctover0046", "Correctover0047"],
         "verified_poc": True,
-        "real_target": "mcp-clickhouse, mcp-server-bigquery (CVSS 7.5)",
+        "real_target": "MCP TypeScript SDK race condition (CVSS 7.1)",
         "consequence": "Internal network access — cloud metadata credential theft",
     },
     "ENV-001": {
         "taxonomy_id": "CRED-001",
-        "cwe": "CWE-200", "cve": "CVE-2026-12957", "cvss": 9.1,
+        "cwe": "CWE-200", "cve": "CVE-2026-12957", "cvss": 7.8,  # 12957 real target = AWS Language Servers; CVSS verified via NVD 2026-08-22 (v3.1 7.8 HIGH / v4.0 8.5)
         "bounty_cases": [],
         "verified_poc": True,
-        "real_target": "CrewAI v1.15.2 (CVSS 9.1, MSRC accepted)",
+        "real_target": "AWS Language Servers (CVSS 7.8, NVD 2026-08-22)",
         "consequence": "Environment variable leak — API key and secret exposure",
     },
     "ENV-002": {
@@ -117,6 +119,22 @@ def _require_pro():
     return True
 
 
+def _gate_scan(validator, is_pro: bool):
+    """Free tier: enforce FREE_DAILY_LIMIT (50 scans/day) before running a scan."""
+    if is_pro:
+        return
+    status = validator.check_license()
+    if not status.get("can_scan", True):
+        limit = status.get("daily_limit", 50)
+        click.echo(
+            f"\n{'━'*55}\n"
+            f"🔒 Free tier daily scan limit reached ({limit}/day).\n"
+            f"   → Upgrade to Pro for unlimited scans: {CTA_URL}\n"
+            f"{'━'*55}\n"
+        )
+        sys.exit(2)
+
+
 def _get_ammo(pattern_id: str) -> dict:
     """Get ammunition data for a runtime guard pattern."""
     return GUARD_AMMO.get(pattern_id, {})
@@ -127,10 +145,16 @@ def _get_ammo(pattern_id: str) -> dict:
 def cli():
     """🛡️ Correctover Runtime Guard — real-time RCE/SSRF/credential leak interception.
 
-    22µs P50 detection. 8 fault patterns cross-referenced with CCS v2.5 taxonomy.
-    Backed by 52 ZDI bounty cases · 8 verified PoCs · 6 CVEs.
+    2011µs P50 detection (local measurement). 8 fault patterns cross-referenced with CCS v2.5 taxonomy.
+    Backed by 52 ZDI bounty cases · 5 verified PoCs · 3 CVEs.
     """
-    pass
+    # Windows GBK terminals crash on non-ASCII output (UnicodeEncodeError).
+    # Force UTF-8 at the CLI output entry; reconfigure() exists in Py3.7+.
+    try:
+        sys.stdout.reconfigure(encoding="utf-8", errors="replace")
+        sys.stderr.reconfigure(encoding="utf-8", errors="replace")
+    except (AttributeError, ValueError):
+        pass
 
 
 @cli.command()
@@ -155,7 +179,7 @@ def start(transport, host, port):
     click.echo(f"  Transport: {BOLD}{transport}{RESET}")
     click.echo(f"  Detection: {GREEN}{LATENCY_P50} P50{RESET}")
     click.echo(f"  Patterns:  {len(FAULT_PATTERNS)} cross-referenced with CCS v2.5 taxonomy")
-    click.echo(f"  {DIM}Backed by 52 ZDI bounty cases · 8 verified PoCs · 6 CVEs{RESET}")
+    click.echo(f"  {DIM}Backed by 52 ZDI bounty cases · 5 verified PoCs · 3 CVEs{RESET}")
     click.echo(f"{BAR}\n")
 
     from .mcp_server import serve
@@ -171,6 +195,7 @@ def diagnose(error_message):
     Pro tier: shows all threats with full repair + evidence.
     """
     is_pro, validator = _get_license_status()
+    _gate_scan(validator, is_pro)
 
     RED = "\033[91m"; YELLOW = "\033[93m"; GREEN = "\033[92m"
     CYAN = "\033[96m"; BOLD = "\033[1m"; DIM = "\033[2m"; RESET = "\033[0m"
@@ -186,11 +211,12 @@ def diagnose(error_message):
         guard._is_pro = False
 
     events = guard.diagnose_error(error_message)
+    validator.record_scan(risks_found=len(events))
 
     click.echo(f"\n{BAR}")
     click.echo(f"  {BOLD}Error Diagnosis Report{RESET}")
     click.echo(f"  Input: {DIM}{error_message[:120]}{'...' if len(error_message) > 120 else ''}{RESET}")
-    click.echo(f"  {DIM}Cross-referenced with CCS v2.5 taxonomy (52 ZDI cases, 8 verified PoCs){RESET}")
+    click.echo(f"  {DIM}Cross-referenced with CCS v2.5 taxonomy (52 ZDI cases, 5 verified PoCs){RESET}")
     click.echo(f"{BAR}\n")
 
     if not events:
@@ -219,7 +245,10 @@ def diagnose(error_message):
             if ammo:
                 evidence_parts = [ammo["cwe"]]
                 if ammo.get("cve"):
-                    evidence_parts.append(f"{ammo['cve']} (CVSS {ammo['cvss']})")
+                    if ammo.get("cvss"):
+                        evidence_parts.append(f"{ammo['cve']} (CVSS {ammo['cvss']})")
+                    else:
+                        evidence_parts.append(f"{ammo['cve']} (CVSS unverified)")
                 click.echo(f"     {CYAN}📎 Maps to: {' | '.join(evidence_parts)}{RESET}")
                 if ammo.get("real_target"):
                     click.echo(f"     {DIM}   Exploited in: {ammo['real_target']}{RESET}")
@@ -244,7 +273,10 @@ def diagnose(error_message):
         if ammo:
             evidence_parts = [ammo["cwe"]]
             if ammo.get("cve"):
-                evidence_parts.append(f"{ammo['cve']} (CVSS {ammo['cvss']})")
+                if ammo.get("cvss"):
+                    evidence_parts.append(f"{ammo['cve']} (CVSS {ammo['cvss']})")
+                else:
+                    evidence_parts.append(f"{ammo['cve']} (CVSS unverified)")
             click.echo(f"     {CYAN}📎 Evidence: {' | '.join(evidence_parts)}{RESET}")
             if ammo.get("real_target"):
                 click.echo(f"     {DIM}   Exploited in: {ammo['real_target']}{RESET}")
@@ -265,7 +297,7 @@ def diagnose(error_message):
             f"   ✓ Repair strategies for each threat\n"
             f"   ✓ Full ZDI bounty case details\n"
             f"   ✓ Real-time blocking ({LATENCY_P50} P50)\n"
-            f"   ✓ Auto-heal (84.1% resolved automatically)\n"
+            f"   ✓ Auto-heal on detected threats\n"
             f"   ✓ Live interception MCP server\n"
             f"{BAR}\n"
             f"   → {CTA_URL}\n"
@@ -287,7 +319,7 @@ def stats():
     click.echo(f"  {BOLD}Runtime Guard Statistics{RESET}")
     click.echo(f"  Detection latency: {GREEN}{LATENCY_P50} P50{RESET}")
     click.echo(f"  Fault patterns:    {len(FAULT_PATTERNS)} cross-referenced with CCS v2.5")
-    click.echo(f"  {DIM}Backed by 52 ZDI bounty cases · 8 verified PoCs · 6 CVEs · max CVSS 9.8{RESET}")
+    click.echo(f"  {DIM}Backed by 52 ZDI bounty cases · 5 verified PoCs · 3 CVEs · max CVSS 9.8{RESET}")
     click.echo(f"{BAR}\n")
 
     from collections import Counter
@@ -303,7 +335,11 @@ def stats():
         ammo = _get_ammo(p.pattern_id)
         click.echo(f"    {sev_color}[{p.severity}]{RESET} {DIM}[{p.pattern_id}]{RESET} {p.description}")
         if ammo:
-            cve_str = f" | {ammo['cve']} (CVSS {ammo['cvss']})" if ammo.get("cve") else ""
+            if ammo.get("cve"):
+                cve_str = (f" | {ammo['cve']} (CVSS {ammo['cvss']})" if ammo.get("cvss")
+                           else f" | {ammo['cve']} (CVSS unverified)")
+            else:
+                cve_str = ""
             click.echo(f"      {CYAN}📎 {ammo['cwe']}{cve_str}{RESET}")
             if ammo.get("real_target"):
                 click.echo(f"      {DIM}   Exploited in: {ammo['real_target']}{RESET}")

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/correctover_agent/guard.py

@@ -1,6 +1,6 @@
 """
 Runtime Guard Engine — RCE/SSRF/Env leak interception via correctover SDK hooks.
-22µs P50 detection latency.
+Real-time pattern detection (P50 latency measured ~2,011µs on local host, 2026-08-22).
 
 v2.0: License enforcement added — core scan requires valid license.
 Free tier gets limited diagnostics (first 2 patterns, no repair).

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/correctover_agent/license.py

@@ -114,6 +114,7 @@ class LicenseValidator:
     STATE_FILE = Path.home() / ".correctover" / "license.json"
     KEY_VERSION = "v2"
 
+    FREE_DAILY_LIMIT = 50           # free-tier scans/interceptions per calendar day
     FREE_FIX_PREVIEW = 2
     FREE_REPORT = False
     FREE_HISTORY = False
@@ -153,22 +154,32 @@ class LicenseValidator:
         self.state["_integrity"] = _state_integrity_hash(content)
         self.STATE_FILE.write_text(json.dumps(self.state, indent=2))
 
+    def _get_day(self) -> str:
+        return time.strftime("%Y-%m-%d")
+
     def _get_product_state(self) -> Dict:
         return self.state["products"].setdefault(self.product, {
             "total_scans": 0,
             "total_risks_found": 0,
             "first_scan": time.time(),
             "last_scan": None,
+            "daily": {},  # {"2026-08-22": count} free-tier per-day scan counter
         })
 
+    def _scans_today(self) -> int:
+        ps = self._get_product_state()
+        return ps.get("daily", {}).get(self._get_day(), 0)
+
     def check_license(self) -> Dict:
-        """Check license status. Returns tier info and capabilities."""
+        """Check license status. Free tier capped at FREE_DAILY_LIMIT scans/day."""
         license_key = self.state.get("license_key") or os.environ.get("CORRECTOVER_LICENSE_KEY")
 
         if license_key and self._verify_license_key(license_key):
             return {
                 "tier": "pro",
                 "can_scan": True,
+                "scans_today": self._scans_today(),
+                "daily_limit": None,
                 "can_fix": True,
                 "can_report": True,
                 "can_heal": True,
@@ -177,9 +188,12 @@ class LicenseValidator:
                 "license_key": license_key[:8] + "..." if license_key else None,
             }
 
+        used = self._scans_today()
         return {
             "tier": "free",
-            "can_scan": True,
+            "can_scan": used < self.FREE_DAILY_LIMIT,
+            "scans_today": used,
+            "daily_limit": self.FREE_DAILY_LIMIT,
             "can_fix": False,
             "can_report": False,
             "can_heal": False,
@@ -193,11 +207,19 @@ class LicenseValidator:
         return self.check_license()["tier"] == "pro"
 
     def record_scan(self, risks_found: int = 0) -> Dict:
-        """Record a scan. Always allowed — this is the hook."""
+        """Record a scan. Free tier is capped at FREE_DAILY_LIMIT scans/day."""
+        status = self.check_license()
+        if status["tier"] == "free" and not status["can_scan"]:
+            raise LicenseExceededError(
+                f"Free tier daily scan limit reached ({self.FREE_DAILY_LIMIT}/day). "
+                f"Upgrade to Pro for unlimited scans."
+            )
         ps = self._get_product_state()
         ps["total_scans"] += 1
         ps["total_risks_found"] += risks_found
         ps["last_scan"] = time.time()
+        daily = ps.setdefault("daily", {})
+        daily[self._get_day()] = daily.get(self._get_day(), 0) + 1
         self._save_state()
 
         history = self.state.setdefault("scan_history", [])

{correctover_runtime_guard-2.1.1 → correctover_runtime_guard-2.1.2}/pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "hatchling.build"
 
 [project]
 name = "correctover-runtime-guard"
-version = "2.1.1"
-description = "Real-time RCE/SSRF/credential leak interception for AI agents — 22µs P50, 99.7% detection rate. MCP Server mode with diagnose_error, fault patterns, repair suggestions. Pro license required for full features."
+version = "2.1.2"
+description = "Real-time RCE/SSRF/credential leak interception for AI agents — MCP Server mode with diagnose_error, fault patterns, repair suggestions. Pro license required for full features."
 readme = "README.md"
 license = { text = "Proprietary Commercial License" }
 authors = [{ name = "Correctover", email = "wangguigui@correctover.com" }]
@@ -35,7 +35,6 @@ Pricing = "https://correctover.com/checkout"
 
 [project.scripts]
 correctover-runtime-guard = "correctover_agent.cli:cli"
-correctover = "correctover_agent.cli:cli"
 
 [tool.hatch.build.targets.wheel]
 packages = ["correctover_agent"]

correctover_runtime_guard-2.1.1/.gitignore

@@ -1,3 +0,0 @@
-
-__pycache__/
-*.pyc

correctover_runtime_guard-2.1.1/.mcp.json

@@ -1,8 +0,0 @@
-{
-  "mcpServers": {
-    "correctover-runtime-guard": {
-      "command": "correctover-runtime-guard",
-      "args": ["start", "--format", "json"]
-    }
-  }
-}

correctover_runtime_guard-2.1.1/LICENSE

@@ -1,17 +0,0 @@
-Copyright (c) 2026 Correctover. All rights reserved.
-
-PROPRIETARY COMMERCIAL LICENSE
-
-This software is proprietary and confidential. Unauthorized copying,
-distribution, modification, or use of this software, via any medium,
-is strictly prohibited without a valid commercial license from Correctover.
-
-Free tier usage is limited to:
-- Diagnostic scanning of up to 2 fault patterns per invocation
-- No repair suggestions, no live interception, no auto-heal
-
-Pro tier features require a valid license key issued by Correctover.
-Purchase at: https://correctover.com/checkout
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND.
-See https://correctover.com/terms for full terms of service.

correctover_runtime_guard-2.1.1/glama.json

@@ -1,15 +0,0 @@
-{
-  "$schema": "https://glama.ai/schema/glama.json",
-  "name": "correctover-runtime-guard",
-  "displayName": "Correctover Runtime Guard — Real-time RCE/SSRF/Env Leak Interception",
-  "description": "Runtime guardrail for AI Agent tool calls. Intercept RCE, SSRF, and environment variable leaks at 22µs P50. MCP Server mode with diagnose_error, get_fault_pattern, and get_repair_suggestion tools.",
-  "homepage": "https://correctover.com/agents/runtime-guard",
-  "category": "Security",
-  "license": "Apache-2.0",
-  "tags": ["runtime", "guard", "rce", "ssrf", "env-leak", "mcp", "agent-security", "correctover", "guardrail"],
-  "pricing": {
-    "model": "freemium",
-    "description": "Enterprise $1,999/mo. Free: 1,500 calls/month (50/day).",
-    "free_tier": "1,500 calls/month (50/day)"
-  }
-}

correctover_runtime_guard-2.1.1/smithery.yaml

@@ -1,5 +0,0 @@
-name: correctover-runtime-guard
-description: Runtime RCE/SSRF/Env leak interception — 22µs P50 via GuardrailProvider
-version: 1.0.0
-runtime: python
-entry: correctover-runtime-guard