Overview

/manipulation-protect is a production command in the Defensive Security category of the Prismatic Platform that activates, configures, and manages manipulation protection defenses across the platform's epistemic systems. When /manipulation-detect identifies manipulation attempts, this command deploys countermeasures that harden the platform against ongoing and future epistemic attacks. The protection system operates at multiple layers -- from individual agent hardening to platform-wide signal validation enforcement.

This command operates under the L3 authority level and is executed by the manipulation-detector agent. It is part of the platform's 216-command slash command registry, built on the AIAD (Autonomous Intelligence Agent Design) standard. The L3 authority level ensures that only authorized operators with security clearance can activate or modify protection defenses, preventing unauthorized weakening of the platform's epistemic integrity.

Manipulation protection in the Prismatic Platform is built on the defense-in-depth principle. No single protection mechanism is considered sufficient. Instead, the system deploys overlapping layers of defense that address each of the five manipulation categories: truth distortion protection (enhanced fact verification), confidence manipulation protection (statistical boundary enforcement), signal poisoning protection (source validation strengthening), drift induction protection (tighter baseline deviation thresholds), and salience hijacking protection (attention distribution monitoring).

The protection system integrates directly with the NABLA epistemic framework. When protection is activated, the NABLA axiom enforcement thresholds are tightened. Signal plurality requirements increase from 2 to 3 independent sources. Confidence scores require additional corroboration. Provenance chains must be complete and verifiable. These tighter requirements reduce the platform's susceptibility to epistemic attacks at the cost of slightly reduced processing speed and increased false-positive rates, a tradeoff that is acceptable during active threat conditions.

The Blue Team develops and validates protection strategies through continuous adversarial testing with the Red Team. The Purple Team synthesizes findings from both teams to close the Red-Blue loop, ensuring that protection defenses are effective against known attack patterns and adaptive to novel techniques.

Architecture

The protection system operates as a multi-layer defense deployment engine with configurable defense profiles and real-time effectiveness monitoring.

Protection Architecture

/manipulation-protect -> Defense Controller -> Layer Deployer -> Monitor
                              |                     |              |
                              v                     v              v
                        Profile Loader         Agent Layer     Effectiveness
                        Rule Compiler          Channel Layer   False Positive
                        Threshold Config       Domain Layer    Performance
                        Policy Enforcer        Platform Layer  Adaptation

Defense Layers

Layer	Scope	Protection Mechanisms	Performance Impact
Agent	Individual agent	Input validation, output verification, behavior constraints	Low (< 5ms per request)
Channel	Data stream	Source authentication, integrity checks, rate limiting	Low (< 10ms per stream)
Domain	Analysis domain	Cross-validation requirements, confidence floor enforcement	Medium (< 50ms per analysis)
Pipeline	Intelligence pipeline	End-to-end provenance verification, plurality enforcement	Medium (< 100ms per pipeline run)
Platform	Entire platform	Global threshold tightening, emergency lockdown capability	Variable

Defense Profiles

Profile	Description	Threshold Tightening	Use Case
Standard	Default protection level	Baseline NABLA thresholds	Normal operations
Elevated	Increased vigilance	+20% threshold tightening	Elevated threat conditions
High	Active defense posture	+50% threshold tightening	Probable manipulation detected
Maximum	Full defensive mode	Maximum threshold enforcement	Confirmed attack in progress
Lockdown	Emergency restriction	All external inputs quarantined	Critical epistemic crisis
Custom	User-defined profile	Configurable per mechanism	Specific threat response

Protection Mechanisms

Mechanism	Category	Description	NABLA Axiom
Enhanced Fact Verification	Truth Distortion	Multi-source fact checking before knowledge base updates	Signal Plurality
Confidence Boundary Enforcement	Confidence Manipulation	Statistical bounds on confidence score changes	Contradiction Preservation
Source Authentication	Signal Poisoning	Cryptographic verification of signal sources	Provenance Mandatory
Drift Threshold Tightening	Drift Induction	Reduced tolerance for behavioral deviation	Time Decay
Attention Distribution Monitoring	Salience Hijacking	Enforced attention balance across targets	Source Independence
Provenance Chain Verification	All Categories	End-to-end provenance for all knowledge updates	Provenance Mandatory
Plurality Enforcement	All Categories	Increased minimum source count for beliefs	Signal Plurality
Quarantine System	All Categories	Isolation of suspicious inputs pending review	Unknown Valid

Usage

# Activate elevated protection
/manipulation-protect activate --profile=elevated

# Activate maximum protection
/manipulation-protect activate --profile=maximum

# Activate protection for specific layer
/manipulation-protect activate --layer=agent --target=sig-osint-commander --profile=high

# Activate protection for specific category
/manipulation-protect activate --category=signal-poisoning --profile=high

# View current protection status
/manipulation-protect status

# View protection for specific agent
/manipulation-protect status --agent=sig-osint-commander

# Deactivate protection (return to standard)
/manipulation-protect deactivate --profile=elevated

# Configure custom protection profile
/manipulation-protect configure --profile=custom-response \
  --plurality-minimum=3 --confidence-floor=0.7 \
  --drift-threshold=0.05 --source-auth=required

# Enable emergency lockdown
/manipulation-protect lockdown --reason="Confirmed epistemic attack on intelligence pipeline"

# Release lockdown
/manipulation-protect release-lockdown --verification="blue-team-clearance-2026-042"

# Quarantine specific input
/manipulation-protect quarantine --source="external-feed-alpha" \
  --reason="Suspected signal poisoning"

# Review quarantined items
/manipulation-protect quarantine --list
/manipulation-protect quarantine --review=QRN-2026-001

# Release quarantined item
/manipulation-protect quarantine --release=QRN-2026-001 --verification="analyst-approval"

# View protection effectiveness metrics
/manipulation-protect metrics --period=7d

# Export protection configuration
/manipulation-protect export --format=json --output=protection-config.json

Options

Option	Type	Default	Description
`action`	string	status	Action: activate, deactivate, status, configure, lockdown, release-lockdown, quarantine, metrics, export
`--profile`	string	standard	Protection profile: standard, elevated, high, maximum, lockdown, custom
`--layer`	string	all	Target layer: agent, channel, domain, pipeline, platform
`--target`	string	all	Specific target within layer
`--category`	string	all	Manipulation category to protect against
`--reason`	string	required for lockdown	Justification for protection activation
`--verification`	string	required for release	Verification code for lockdown release
`--source`	string	none	Source to quarantine
`--plurality-minimum`	integer	2	Minimum source count for custom profiles
`--confidence-floor`	float	0.5	Minimum confidence for custom profiles
`--drift-threshold`	float	0.1	Maximum drift tolerance for custom profiles
`--source-auth`	string	recommended	Source authentication: optional, recommended, required
`--period`	string	24h	Metrics reporting period
`--format`	string	text	Output format: text, json, markdown
`--output`	string	stdout	Output file path

Execution Flow

Profile Resolution: The requested protection profile is resolved, loading all associated threshold values, mechanism configurations, and enforcement policies. Custom profiles are compiled from provided parameters.
Current State Assessment: The current protection state is assessed to determine what changes are needed. Upgrading from standard to elevated requires different actions than upgrading from elevated to maximum.
Layer Deployment: Protection mechanisms are deployed across the specified layers. Each layer receives its appropriate mechanism configurations. Deployment is atomic -- either all mechanisms deploy successfully or the entire activation is rolled back.
NABLA Threshold Adjustment: The NABLA framework's enforcement thresholds are adjusted according to the protection profile. Plurality requirements, confidence floors, provenance depth, and time decay parameters are all updated.
Agent Hardening: Individual agents receive updated behavioral constraints. Input validation rules are tightened. Output verification requirements are increased. Cross-validation frequencies are adjusted.
Channel Protection: Intelligence channels receive enhanced source authentication requirements, integrity checking, and rate limiting. Suspicious sources may be automatically quarantined.
Monitoring Activation: Enhanced monitoring is activated alongside protection mechanisms. The monitoring system tracks protection effectiveness, false positive rates, and performance impact in real-time.
Effectiveness Tracking: All protection activations are tracked with timestamps, triggering events, and effectiveness metrics. This data feeds the protection optimization loop and provides evidence for protection decisions.
Notification: Relevant operators and agents are notified of protection state changes. The /manipulation-dashboard is updated to reflect the new protection posture.

Integration Points

Component	Relationship	Details
Prismatic Agents	Executed by `manipulation-detector`	Defense deployment and monitoring
/manipulation-detect	Detection triggers	Detections trigger protection activation
/manipulation-dashboard	Status display	Protection status shown in dashboard
/manipulation-techniques	Countermeasure reference	Technique-specific countermeasure deployment
Color-Team	Security framework	Blue Team defense strategies
NABLA Framework	Threshold management	Axiom enforcement level adjustment
Quality Gates	Quality integration	Protection integrated into quality checks
Telemetry	Event tracking	Protection events logged
/emergency	Crisis coordination	Lockdown coordinated with emergency protocol

Best Practices

Escalate protection gradually. Start with elevated protection and increase to high or maximum only if the threat persists or escalates. Maximum protection significantly impacts processing speed and increases false positives.

Monitor false positive rates. Protection mechanisms that are too aggressive create operational friction. Track false positive rates through /manipulation-protect metrics and adjust thresholds if false positives exceed 5% of total checks.

Use custom profiles for targeted threats. When the specific manipulation technique is identified, create a custom profile that targets that technique aggressively while leaving other mechanisms at standard levels. This minimizes collateral impact.

Document lockdown reasons thoroughly. Lockdowns are the most disruptive protection action. Clear, specific documentation of the triggering event and the expected conditions for release ensures that lockdowns are released promptly when safe.

Review quarantine regularly. Quarantined inputs accumulate if not reviewed. Establish a regular review cadence (daily during active threats, weekly during normal operations) to prevent legitimate data from being indefinitely blocked.

Test protection effectiveness with Red Team. Periodically activate protection profiles and run Red Team scenarios to validate that defenses are effective. Protection mechanisms that have not been tested against adversarial scenarios provide false confidence.

Error Handling

Error	Cause	Resolution
`authority_insufficient`	Operator lacks L3 authority	Request authority elevation
`profile_not_found`	Specified protection profile does not exist	Use built-in profile or create custom one
`lockdown_requires_reason`	Lockdown requested without justification	Provide `--reason` parameter
`release_requires_verification`	Lockdown release without verification code	Obtain verification from Blue Team commander
`deployment_failed`	Protection mechanism failed to deploy	Check system health, review deployment logs
`threshold_conflict`	Custom thresholds conflict with system requirements	Adjust custom thresholds to be within valid ranges
`quarantine_full`	Quarantine capacity exceeded	Review and release cleared items
`target_not_found`	Specified protection target does not exist	Verify target name with system inventory

Advanced Usage

Automated Response Chains

Configure automated protection responses triggered by detection events.

# Auto-activate elevated on YELLOW threat
/manipulation-protect configure --auto-response \
  --trigger-level=elevated --profile=elevated

# Auto-lockdown on CRITICAL with multiple confirmations
/manipulation-protect configure --auto-response \
  --trigger-level=critical --profile=lockdown \
  --require-confirmations=2

# Auto-quarantine untrusted sources
/manipulation-protect configure --auto-quarantine \
  --source-trust-threshold=0.3

Protection Testing

Validate protection effectiveness against known attack patterns.

# Run protection validation suite
/manipulation-protect test --scenarios=all --profile=elevated

# Test specific mechanism
/manipulation-protect test --mechanism=source-authentication --attack=signal-poisoning

# Benchmark protection performance impact
/manipulation-protect benchmark --profile=maximum --duration=60m

Cross-Platform Protection

Coordinate protection across connected platform instances.

# Synchronize protection state across instances
/manipulation-protect sync --instances="prod,staging" --profile=elevated

# Propagate lockdown to all instances
/manipulation-protect lockdown --propagate=all --reason="Coordinated attack detected"

Doctrine Compliance

All commands operate under the NO MERCY, NO DOUBTS doctrine:

NO MERCY: Zero tolerance for incomplete execution or quality violations. Protection mechanisms deploy fully or not at all -- no partial deployments. Lockdowns execute immediately without negotiation.
NO DOUBTS: Full investigation before action, evidence-based results. Protection decisions are grounded in detection evidence. Effectiveness is measured, not assumed. False positive rates are tracked and managed.

/manipulation-detect - Detect manipulation attempts using epistemic analysis
/manipulation-techniques - View manipulation technique taxonomy and counter-measures
/manipulation-dashboard - Manipulation detection dashboard with threat indicators
/emergency - Emergency response and crisis management activation
/archer-supreme - Supreme authority activation for platform-wide operations
/dark-ops - NABLA structural crisis detection and dark operations analysis

Connect & Contribute

Created by Tomáš Korcak (korczis) | Open Source under GHL

GitHub | GitLab | LinkedIn | Contact
Developer Portal | Architecture | Meet the Creator

/manipulation-protect