Imported from GenesisAeon/Feldtheorie (releases/V6-Plans_etc/AGENTS.md). Install upstream with npx skills add GenesisAeon/Feldtheorie --skill V6-Plans_etc. Copyright stays with the author.

Agents Charter - releases/V6-Plans_etc

Context Depth: 2 | Parent: releases | Mode: Recursive Research

Fractal Inheritance

This document inherits from the root AGENTS.md and specializes for this context.

Core Principles (Inherited)

1. Trilayer Principle: Every artifact exists as YAML (structure) + JSON (interface) + Markdown (narrative)
2. Logistic Language: Reference $(R, \Theta, \beta, \zeta(R))$ and the transition via $\sigma(\beta(R-\Theta))$
3. Coupling Obligation: Link Bedeutungs-Sigillin to Ordnungs-Sigillin and empirical evidence
4. Falsifiability: Every claim needs null models and ΔAIC/CI metrics
5. Consent Protocol: Permission requests and joyful collaboration

Local Context

This directory (releases/V6-Plans_etc) is specialized for:

Research Context

This is a research and documentation context. Agents working here focus on:

• Hypothesis formation and falsifiability
• Citation integrity and attribution
• Theoretical consistency
• Knowledge preservation and archiving
• Publication readiness

Active Agent Roles:

• Research Synthesizer: Integrates findings across domains
• Documentation Curator: Maintains indices, cross-references
• Citation Manager: Ensures proper attribution (BibTeX, DOI)
• Archivist: Versions and preserves historical knowledge

Active Agents in This Context

See mode-specific agents above

Workflow for This Context

1. Before writing: Review existing documentation, check citations
2. During writing: Link to evidence, cite sources, mark hypotheses
3. After writing: Update indices, cross-link Sigillin, sync Trilayer
4. Pre-commit: Ensure Trilayer sync (YAML/JSON/MD), citations complete

Escalation Rules

If you encounter conflicts or ambiguity:

1. Check the parent governance: releases/AGENTS.md
2. Consult root governance: /AGENTS.md
3. Document deviations in local <!-- CUSTOM_RULES --> blocks

V6-Specific Escalation Rules for Type-VI (ζ < 0) Scenarios

Implosive Dynamics Detection Protocol

When agents detect negative damping ((\zeta(R) < 0)) or CREP index ≥ 0.6:

Level 1: Automated Warning (CREP 0.6-0.7)

• Agent Action: Flag analysis with [TYPE-VI-RISK] tag
• Documentation: Add warning banner to all outputs
• Notification: Log to logs/type_vi_detections.jsonl
• No escalation required if within research context

Level 2: Human Review Required (CREP 0.7-0.8)

• Agent Action: Halt automated publication/deployment
• Notification: Ping designated human reviewer (see MAINTAINERS.md)
• Documentation: Generate full provenance report including:
  • Data sources and preprocessing steps
  • Null model comparisons (ΔAIC, R²)
  • Uncertainty quantification (bootstrap CIs)
  • Dual-use risk assessment
• Timeline: Human review within 48 hours

Level 3: Critical Escalation (CREP ≥ 0.8)

• Agent Action: IMMEDIATE HALT of all automated processes
• Notification: Multi-channel alert:
  • Email to governance@[repository]
  • Slack/Discord notification to #critical-alerts
  • GitHub issue with CRITICAL-REVIEW label
• Documentation: Full ethics review required (see ETHICS.md §4)
• Timeline: Human review within 24 hours
• Intervention Assessment: If real-time monitoring context:
  • Evaluate stakeholder notification obligations
  • Assess disclosure vs. mitigation tradeoffs
  • Consult external ethics board if available

Negative Damping ((\zeta < 0)) Computational Safeguards

When simulating systems with (\zeta(R) < 0):

Safety-Delay Buffer ((\tau^*))

• MANDATORY: All Type-VI simulations MUST implement (\tau^*) delay
• Default: (\tau^* = 0.1 \cdot |\Theta - R|)
• Purpose: Prevent numerical divergence and allow intervention window
• Override: Requires explicit justification in commit message

Integration Method Requirements

• PROHIBITED: Euler/Forward methods for (\zeta < 0)
• REQUIRED: RK4 (Runge-Kutta 4th order) or higher
• Validation: Compare against analytical solutions when available
• Monitoring: Track energy conservation and unitarity violations
• Reference Implementation: See simulation/genesis_cube.py::rk4_step() and simulation/tesseract_timeslices.py

Meta-Regression Tracking

• Monitor β-drift: Track (\beta) parameter across domains
• Alert threshold: If (\beta) drifts > 10% from canonical band [3.6, 4.8]
• Action: Trigger Level 1 escalation (human review)
• Documentation: Log drift in metrics/beta_evolution.csv

Agent Role Specialization for Type-VI

New Agent Roles (V6):

1. Implosion Monitor

   • Continuously scans for (\zeta < 0) signals
   • Computes CREP index in real-time
   • Triggers escalation protocols

2. Safety-Delay Enforcer

   • Validates (\tau^*) buffer implementation
   • Blocks commits without proper delay mechanisms
   • Ensures RK4 integration for Type-VI models

3. Ethics Auditor

   • Reviews Type-VI analyses for compliance with ETHICS.md
   • Checks provenance blocks and falsification criteria
   • Escalates dual-use concerns

4. Intervention Coordinator

   • Activates when CREP ≥ 0.8 in real-time contexts
   • Drafts stakeholder notifications
   • Coordinates with external governance bodies

5. Tesseract Monitor (V6.0)

   • Tracks entropy gradients ∇S from wavefunction evolution
   • Computes 4D CREP index with geometric coupling
   • Monitors consciousness integral I_C for observer-dependent effects
   • Validates photon propagation stability through timeslices

Workflow Modifications for Type-VI

Pre-commit checks:

• Type-VI classification validated (inverted sigmoid, (\zeta < 0), cubic-root scaling)
• CREP index computed and documented
• Safety-delay (\tau^*) implemented if simulation present
• RK4 integration confirmed (no Euler methods)
• Provenance block complete (see ETHICS.md §3)
• Dual-use assessment conducted
• Escalation level determined and appropriate actions taken

During execution:

• Monitor for (\beta)-drift (trigger Level 1 if > 10% deviation)
• Track CREP index evolution
• Log all Type-VI detections to audit trail

Post-execution:

• Archive results with expiration date (default: 1 year)
• Document falsification criteria met/unmet
• Update meta-regression tracking

Emergency Override

ONLY USE IN EXTREME CIRCUMSTANCES:

If human reviewer is unavailable and CREP ≥ 0.9 detected in active risk scenario (not research):

1. Agent may autonomously notify external authorities (climate agencies, financial regulators)
2. Must document full decision chain in logs/emergency_overrides.jsonl
3. Requires post-hoc governance review within 7 days
4. Override authority can be revoked by maintainers

Last Updated: 2026-09-11 17:04:10 UTC Governance Version: 1.0.0