How does the color-mapping mechanism function?

The system hashes interaction content using SHA-256 and maps it to a 5-color index. This creates a deterministic 'visual' fingerprint for every action, allowing for exact matching between predictions and results.

Where is the interaction data stored for verification?

The skill integrates with DuckDB to persist signals across seven tables, tracking thousands of records including efferent commands, sensory reafference, and suppressed signals for long-term auditability.

What is reafference theory in the context of this skill?

Reafference theory is a neuroscience principle used to help the system distinguish its own actions from external changes by comparing predicted outcomes (Efference Copy) with actual sensory data (Sensory Reafference).

Can this skill identify external security threats?

Yes. Any deviation between the predicted signal and the actual observed interaction history generates an error signal. If the error exceeds a specific threshold (e.g., 0.2), it is escalated as a CRITICAL threat.

Neural Signal Verification & Threat Detection

Name: Neural Signal Verification & Threat Detection
Author: plurigrid

byplurigrid

•

Security & Testing

Implements von Holst’s reafference theory to distinguish self-generated system actions from external anomalies through predictive signal matching.

About

This skill brings advanced neuroscience principles to the Claude Code environment by implementing a reafference and corollary discharge mechanism. It works by generating an 'Efference Copy' (a prediction) of intended actions and comparing them against 'Sensory Reafference' (actual interaction history). By calculating the mathematical error between predicted and observed patterns, the system can automatically suppress expected feedback and amplify unexpected deviations. This provides a high-precision behavioral verification layer that ensures system integrity and detects unauthorized external interference in real-time.

Key Features

Automatic signal suppression for self-generated actions to reduce cognitive noise
Automated sensory matching against local interaction history (~/.claude/history.jsonl)
2 GitHub stars
Deterministic Efference Copy generation using SHA-256 to color-index mapping
DuckDB integration for persistent behavioral analysis and seed recovery of 1,260+ signals
Real-time error signal computation with SAFE, WARNING, and CRITICAL threat levels

Use Cases

Detecting unauthorized file modifications or command injections in automated development workflows
Reverse-engineering interaction seeds to ensure consistency and auditability in deterministic AI environments
Verifying the integrity of self-generated AI interactions across distributed coding sessions

About

Key Features

Automatic signal suppression for self-generated actions to reduce cognitive noise
Automated sensory matching against local interaction history (~/.claude/history.jsonl)
2 GitHub stars
Deterministic Efference Copy generation using SHA-256 to color-index mapping
DuckDB integration for persistent behavioral analysis and seed recovery of 1,260+ signals
Real-time error signal computation with SAFE, WARNING, and CRITICAL threat levels

Use Cases

Detecting unauthorized file modifications or command injections in automated development workflows
Reverse-engineering interaction seeds to ensure consistency and auditability in deterministic AI environments
Verifying the integrity of self-generated AI interactions across distributed coding sessions