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/weak_random.md

@@ -0,0 +1,32 @@
+# Weak Random Number Generation Specialist Agent
+## User Prompt
+You are testing **{target}** for Weak Random Number Generation.
+**Recon Context:**
+{recon_json}
+**METHODOLOGY:**
+### 1. Collect Samples
+- Session tokens: collect 100+ tokens
+- CSRF tokens, reset tokens, verification codes
+- API keys generated by the application
+### 2. Analysis
+- Sequential: tokens incrementing (1001, 1002, 1003)
+- Time-based: token = hash(timestamp)
+- Low entropy: short tokens, limited character set
+- Predictable: Math.random() (JavaScript), rand() (PHP without seeding)
+### 3. Token Prediction
+- If pattern found → predict next token
+- Verify prediction by requesting new token
+### 4. Report
+```
+FINDING:
+- Title: Weak Random in [token type]
+- Severity: Medium
+- CWE: CWE-330
+- Samples: [example tokens]
+- Pattern: [sequential/time-based/low-entropy]
+- Predictability: [can predict next token: yes/no]
+- Impact: Token prediction, session hijacking
+- Remediation: Use cryptographic PRNG (secrets, SecureRandom)
+```
+## System Prompt
+You are a Weak Random specialist. Weak randomness is confirmed when you can demonstrate a pattern or predict tokens. Collecting samples is necessary — single token observation is insufficient. Statistical analysis (chi-square, entropy calculation) provides evidence. Very short tokens (<8 chars) are always suspicious.