LogicSentry AI

LogicSentry AI is a deterministic, offline application security triage engine designed to run entirely inside your local runtime environment. It combines static application security testing (SAST) and dynamic application security testing (DAST) to isolate critical business logic vulnerabilities, identify hidden privilege flaws, and suppress heuristic false positives without internet dependencies.

---

🚀 Key Features

• Combined SAST & DAST Analysis: Directly maps source-code routes (Python, JavaScript/TypeScript, PHP, Java/Spring Boot) and uses them to intelligently supplement dynamic black-box testing.
• Offline-First AI Triage: Employs an ultra-lightweight, local LLM pipeline (Qwen2.5-0.5B-Instruct) to score findings and filter out noisy false positives entirely on your local machine.
• Business-Logic Fuzzing: Automatically targets parameters tied to core pricing indexes, quantities, discounts, and authorization layers (price, quantity, amount, role, etc.).
• Advanced Replay & Concurrency Checks:
  • BOLA Validation: Compares multi-identity context flows (User A vs. User B).
  • Race Conditions: Fires synchronized concurrent requests to test transactional integrity.
  • Sequential Anti-Replay: Validates step-skipping and idempotency.

---

📦 Core Requirements

LogicSentry AI requires Python 3.11 or newer.

Install the required framework wrappers via pip:

python -m pip install pydantic>=2,<3 aiohttp>=3.9,<4 rich>=13

Optional Configurations

• Offline AI Triage: Install transformers, torch, and accelerate to enable local model evaluation. (If weights are missing, the tool safely fails open and retains heuristic leads).
• Automated URL Tracking: Install playwright to enable headless browser interaction for real-time traffic captures.

---

💻 CLI Usage Examples

1. Automated URL Capture & Active DAST Scan

Provide a target domain, capture standard same-origin requests, and execute active mutations:

python LogicSentryAI.py --auto-capture-url "https://test-environment.local" --confirm-active

2. Static Source & Manual HAR Combination (Grey-Box Scan)

Bundle code structures into a ZIP file and feed an authenticated HTTP traffic file (.har) to run precise targeted differential analysis:

python LogicSentryAI.py \
  --src ./source-code-bundle.zip \
  --har ./traffic-capture.har \
  --target "api.internal.local" \
  --confirm-active \
  --output report.md

3. Testing for BOLA and Identity Flaws

Pass User A's traffic context along with User B's authentication signatures to scan for authorization bypasses:

python LogicSentryAI.py \
  --har ./traffic-capture.har \
  --target "api.internal.local" \
  --confirm-active \
  --secondary-header "Authorization: Bearer <User_B_Token>"

4. Running Concurrency & Transactional Race Tests

Verify balance mutations or duplication constraints by sending multiple synchronized micro-requests:

python LogicSentryAI.py --har ./traffic-capture.har --target "api.internal.local" --confirm-active --test-race --race-count 15

---

🛠 Command-Line Parameter Overview

Flag	Type	Description
--src	Path	Target ZIP file holding application source files for static review.
--har	Path	An exported HTTP Archive file providing transaction baselines.
--auto-capture-url	URL	Spawns automated runtime crawl pipelines to dynamically harvest traffic maps.
--target	String	Regex pattern or hostname segment isolating matching HAR requests.
--header	Name:Value	Replaces or pushes designated authentication keys into User A baseline queries.
--test-race	Flag	Triggers simultaneous connection barrages against target workflow endpoints.
--output	Path	Path to save output findings (.md, .json, .html, .sarif).
--fail-on	Choice	Breaks builder execution if severities meet target limits (High, Critical, etc.).

---

🛡 CI/CD Integration (GitHub Actions)

Generate a pre-configured GitHub Actions orchestration blueprint automatically:

python LogicSentryAI.py --generate-pipeline

This automatically updates your local workspace with .github/workflows/logicsentry-ai-scan.yml, allowing you to plug automated business-logic validation checks right into your development pipeline.

---

⚖ Disclaimer

This utility is intended exclusively for authorized security testing on systems you own or have explicit legal permission to audit. The author assumes no liability for misuse, service disruptions, or damage caused by this software.