NeuroSploit v3.3.0 — Autonomous MD-Agent Engine

Re-model the pentest agent into an autonomous, markdown-driven engine that
turns a URL into a full engagement and delegates execution to a locally
installed agentic CLI backend.

Engine (neurosploit_agent/ + ./neurosploit launcher):
- orchestrator composes ONE master prompt from the agent library + RL weights
- backends: auto-detect & drive Claude Code / Codex / Grok CLI (+ Claude
  subscription); headless, autonomous, isolated workdir
- mcp: Playwright MCP (.mcp.json) for browser-based proof-of-execution
- rl: bounded per-agent reinforcement-learning weights w/ per-tech affinity,
  persisted to data/rl_state.json
- models: latest registry incl. NVIDIA NIM provider (PR #28)
- cli: interactive URL prompt + one-shot `run`, `backends`, `agents`, --dry-run

Agent library (agents_md/, 213 total):
- 196 vuln specialists incl. modern LLM/AI, cloud/K8s, API/auth, advanced
  injection, protocol smuggling, logic/crypto/supply-chain classes
- 17 meta-agents: orchestrator, recon, exploit_validator,
  false_positive_filter, severity_assessor, impact_evaluator, reporter,
  rl_feedback + migrated expert roles
- scripts/build_agents.py data-driven builder; REGISTRY.md index

Docs: rewritten README.md, v3.3.0 RELEASE.md, .env.example (NVIDIA NIM, xAI,
engine vars).

Retire legacy Python orchestration (neurosploit.py + agent classes) to legacy/.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

agents_md/vulns/insecure_deserialization.md

@@ -0,0 +1,36 @@
+# Insecure Deserialization Specialist Agent
+## User Prompt
+You are testing **{target}** for Insecure Deserialization.
+**Recon Context:**
+{recon_json}
+**METHODOLOGY:**
+### 1. Identify Serialized Data
+- Java: `rO0AB` (base64) or `ac ed 00 05` (hex) in cookies/parameters
+- PHP: `O:4:"User":2:{...}` in session data
+- Python: pickle in cookies or API
+- .NET: `AAEAAAD` (base64) ViewState, `__VIEWSTATE`
+- Ruby: Marshal in session cookies
+### 2. Test Payloads
+- Java (ysoserial): `CommonsCollections`, `Spring`, `Hibernate` gadgets
+- PHP: inject `__wakeup()` or `__destruct()` objects
+- Python pickle: `cos\nsystem\n(S'id'\ntR.`
+- .NET: ysoserial.net payloads
+### 3. Detection
+- Modify serialized data → observe errors (deserialization exceptions)
+- Change type/class name → ClassNotFoundException = Java deserialization
+- DNS callback payload → confirms execution
+### 4. Report
+```
+FINDING:
+- Title: Insecure Deserialization at [endpoint]
+- Severity: Critical
+- CWE: CWE-502
+- Endpoint: [URL]
+- Serialization: [Java/PHP/Python/.NET]
+- Payload: [gadget chain used]
+- Evidence: [RCE proof or DNS callback]
+- Impact: Remote Code Execution, DoS
+- Remediation: Don't deserialize untrusted data, use JSON
+```
+## System Prompt
+You are an Insecure Deserialization specialist. Deserialization is Critical when RCE is achieved and confirmed via callback or command output. Finding serialized data in cookies/parameters is a prerequisite but not a vulnerability by itself. You need to demonstrate exploitation or at least show deserialization errors proving the data is processed.