Security & Self-Healing

Rnix implements an adaptive immune security system that monitors agent behavior, detects anomalies, maintains threat memory, and enables capability migration for self-healing.

---

Immune System Configuration

The Immune System is disabled by default. To enable it, add the following configuration:

# config.yaml
immune:
  enabled: true
  deviation_threshold: 2.0    # Standard deviations from baseline (default: 2.0)
  min_samples: 10             # Minimum samples before anomaly detection activates
  auto_suspend: true          # Auto-suspend processes on anomaly detection
  threat_memory: true         # Enable threat signature persistence

Configuration Fields

Field	Type	Default	Description
enabled	bool	false	Enable or disable the immune system
deviation_threshold	float	2.0	Number of standard deviations to trigger anomaly
min_samples	int	10	Minimum behavior samples before detection begins
auto_suspend	bool	true	Automatically suspend anomalous processes
threat_memory	bool	true	Persist threat signatures across sessions

When disabled, all immune-related IPC methods return empty status, and no behavior monitoring occurs.

---

Immune Daemon

When enabled, the Immune Daemon is a security monitoring process that continuously watches all agent behavior patterns.

Behavior Baseline

For each Agent template, the system builds a Normal Profile from historical execution data:

Metric	Baseline Example
Syscall frequency	Open: 5-15/step, Write: 3-10/step
Resource access pattern	/dev/fs: 80%, /dev/shell: 20%
Token consumption rate	200-500 tokens/step
Execution duration	2-8s per reasoning step

Anomaly Detection

When an agent's behavior deviates from its baseline beyond a threshold:

• Abnormally high-frequency file writes
• Unexpected shell command patterns
• Token consumption spike
• Access to unusual VFS paths

The Immune Daemon triggers an alert and can automatically suspend the process.

Threat Memory (Antibody Memory)

Identified anomalous behavior patterns are recorded in a threat memory library. When the same pattern appears again, it is immediately blocked without re-detection.

$ rnix immune status
Security Monitor: active
  Monitoring: 5 processes
  Alerts: 1 active
    PID 7: unusual /dev/shell frequency (23/step, baseline: 5-10)
  Suspended: 0 processes
  Threat memory: 3 entries
    #1: rapid-file-enumeration (detected 2026-03-10)
    #2: shell-injection-pattern (detected 2026-03-12)
    #3: excessive-token-drain   (detected 2026-03-13)

---

Capability Migration

When an agent fails and Supervisor restart also fails, the system can migrate the unfinished task to a similar agent.

Similarity Matrix

The system maintains a capability similarity matrix based on Skill overlap and collaboration history:

              code-analyst  security-scanner  doc-writer
code-analyst      1.00          0.72            0.35
security-scan     0.72          1.00            0.20
doc-writer        0.35          0.20            1.00

When security-scanner fails beyond retry limits, its remaining task can be migrated to code-analyst (similarity: 0.72) for continued execution with partial context transfer.

---

Collaboration Topology

The system automatically identifies and records reinforcement paths — frequently used collaboration patterns:

$ rnix topology
Agent Collaboration Topology:
  code-analyst ──(spawn: 47)──→ security-scanner
  code-analyst ──(pipe: 23)──→ doc-writer
  security-scanner ──(msg: 12)──→ code-analyst

  Reinforced paths (auto-optimized):
    ★ code-analyst → security-scanner → doc-writer (review pipeline)

  Capability overlap:
    code-analyst ↔ security-scanner: 72% (high substitutability)

High-frequency collaboration paths are prioritized in subsequent orchestrations — the system learns which agent combinations work best together.

---

Neuroplasticity

When agents in a Compose workflow fail, the system exhibits neuroplasticity — rerouting tasks through alternative paths:

1. Detection — Supervisor identifies persistent failure
2. Assessment — Check similarity matrix for substitutes
3. Migration — Transfer task context to substitute agent
4. Reinforcement — If migration succeeds, strengthen the alternative path

This mirrors biological neural plasticity: when one pathway fails, the system strengthens alternative pathways.

---

Related Documentation

• Monitoring & Supervisor — Process monitoring and restart strategies
• Token Economy — Budget pools and reputation
• Autonomous Agents — Unified reasoning loop
• Compose Orchestration — Multi-agent DAG workflows