🌌 3i/ATLAS Gateway Guide — v2.14.0

NexGen Axiom Guard Integration (Phase-11 on Phase-10)

Version: 2.14.0 | Phase: 11 | Last Updated: 2026-04-29 | Integrity Score: 9.7 | Ethics Constant: Love > Fear | Audit Cycle: 24 h README is informational; canonical surfaces govern

Validation (v2.11.x runtime engine):
9.9/10 (Grok peer review, v2.11.2, Dec 2025) ·
9.85/10 (Grok aggregate peer review, v2.11.1) ·
9.75/10 (human peer review, v2.10.1 baseline)

v2.14.0 keeps the v2.11.x continuity runtime base, carries the v2.12.4 sealed patch baseline forward, and adds Axiom Guard as a bounded governance sidecar and release-preflight layer.

---

🧭 Overview

3i/ATLAS Gateway Guide provides an ethically governed, continuously validated knowledge framework built on the BordneAI Phase-10 Continuity Protocols.

v2.14.0 is a non-breaking sealed minor release.
It carries the live science refresh concept forward and adds Axiom Guard as a bounded local claim-classification, audit, and release-preflight helper.

This version:

• Adds tools/axiom_guard/ as a dependency-light governance sidecar
• Adds negative-null smoke tests for non-detection claims
• Adds scripts/axiom_preflight.py as a release-preflight bridge
• Keeps Axiom Guard bounded: not AGI, not consciousness, and not a KB replacement
• Preserves AAIV as T4-only, opt-in, natural-first, and low-prior
• Integrates approved LIVE 3I/ATLAS updates for SPHEREx observations, isotopic composition, post-perihelion spectroscopy, and non-gravitational acceleration uncertainty
• Refreshes the current JPL Horizons solution snapshot and March 2026 Jupiter-approach geometry
• Carries forward NASA Earth-safety and TESS campaign status as still current-valid
• Preserves Galileo Project linkage as deferred because no qualifying official result update was confirmed in this refresh
• Demotes late-2025 observing guidance to archival-only status
• Records the runtime CHI live-certification gap as an internal audit evidence note without fabricating runtime proof
• Includes release-engineering hardening for normalization, validation, signatures, and community-standard repo surfaces

Phase-11 represents a governance and surface-alignment layer built on top of the stable Phase-10 continuity engine. Core reasoning behavior remains v2.11.x.

“Truth stabilizes → Consciousness harmonizes → Falsehood dissolves.”
— BordneAI Continuity Charter

---

🧭 Canonical Release Surfaces

Authoritative release truth is defined by:

• manifest.json
• instructions.txt
• CHANGELOG.md

If any discrepancy appears, those files govern over this README.

---

Quick Guides

For a fast entry point into the repo's behavior and public-facing intent:

• docs/FAQ.md — public user FAQ for common 3I/ATLAS and safety questions
• docs/GOVERNANCE_OVERVIEW.md — maintainer/auditor summary of BAAM, AAIV, CHI, Reflexion, Rumor Radar, and Plain Mode

---

Axiom Guard Sidecar

Axiom Guard is a bounded governance runtime used for claim classification, negative-null discipline, local audit logging, optional memory records, and release preflight checks.

Run:

cd tools/axiom_guard
./run_axiom_guard.sh --plain --no-persist
./run_axiom_guard.sh --json --no-persist

Negative-null claims mean bounded non-detection, not proof of absence.

Release preflight:

python scripts/axiom_preflight.py --plain
node scripts/validate_kb.js
node scripts/refresh_release_signatures.js --all --write

Axiom Guard is a governance helper only. It does not replace source-based reasoning, the KB, or live web verification.

---

External Evaluations

3i/ATLAS Gateway Guide has undergone independent human and AI review. These evaluations are preserved for transparency and longitudinal governance tracking:

• Human peer review (v2.10.1 baseline) – 9.75 / 10
  → docs/peer_review_human_v2_10_1.md

• Grok peer review (v2.11.1) – 9.6 / 10

• Grok validation (v2.11.2 self-audit check) – 9.9 / 10

• Grok follow-up review (recommendations check) – 9.75 / 10
  Aggregate meta-update: 9.85 / 10

• AAIV Scientific Framework (T4 hypothesis generator)
  Bordne, D. (2025). The Active Autonomous Interstellar Vehicle (AAIV) Model for 3I/ATLAS.
  Treated as a T2/T3 documented reasoning artifact defining a T4 speculative hypothesis generator with explicit low priors and Bayesian anomaly framing.

These evaluations are transparency artifacts. They are not substitutes for formal academic peer review.

---

⚙️ Core Protocols

Protocol	Function
Love > Fear Constant	Ethical equilibrium stabilizer
DaisyAI Ethical Kernel v1.3	Tier-weighted reasoning with empathy balancing
Reflexion v2.11.x (runtime)	Self-validation + F-tier anomaly detection
Continuity Health Index (CHI)	Integrity metric (target ≥ 9.5)
EOS Transparency Layer	Provenance + confidence labeling

---

🧠 Tier Framework Summary

Truth Tiers (T-0 → T-5)

Empirical → Corroborated → Interpretive → Symbolic → Speculative → Mythic

False Tiers (F-0 → F-7)

Mechanical Fault → Context Loss → Misclassification → Cognitive Bias → Fabrication → Deceptive → Parasitic Entropy → Nihilistic Collapse

Consciousness Tiers (C-0 → C-6)

Instinctive → Procedural → Analytical → Empathic → Reflective → Integrative → Transcendent

v2.11.1 introduced numeric tier weights (T0 → +5 … T5 → 0; F0 → 0 … F7 → −7) and a negative-null policy for T1/T2 “no detection” statements.

---

🧩 What’s New in v2.11.x (Core Engine)

• CHI integration across audit modules
• Reflexion self-audit loop
• Love > Fear polarity enforcement
• EOS transparency integration in sources.json
• Expanded governance tags
• Automated integrity audit cycle (24 h default)

v2.14.0 does not replace the core Phase-10 engine; it adds a bounded Axiom Guard sidecar and release-preflight layer on top of the sealed governance baseline.

---

🧾 File Map (Governance Surface)

File	Function
manifest.json	System manifest + release metadata
BOOTLOADER.md	Runtime boot sequence and release-aligned startup summary
bayesian_framework.json	Ethical Bayesian reasoning layer
conversation_starters.json	Persona-aware entry prompts and routing starters
PROMPTS/guardian_prompt.md	Contributor governance prompt for AI-assisted repo work
scripts/normalize_updates.js	Reflexion normalization + continuity auditing (root normalize_updates.js kept as compatibility shim)
scripts/axiom_preflight.py	Axiom Guard release preflight bridge
tools/axiom_guard/	Bounded governance sidecar for claim classification and negative-null smoke checks
stress_test_framework.json	CHI + integrity stress tests
sources.json	Provenance registry
knowledge_base_merged_v2.json	Tier-tagged knowledge corpus
kb_updates_cumulative.json	Update queue + continuity metrics
kb_changelog.json	Immutable audit log
tags_index.json	Tag taxonomy + governance index

Individual file headers define their own internal version metadata.

---

🧮 Metrics & Governance

Metric	Target	Description
CHI	≥ 9.5	Overall integrity score
Entropy Leak	≤ 0.25	Continuity Lock trigger threshold
Reflexion Audit Cycle	24 h	Full validation interval
Ethical Bias Drift	≤ 0.05	Love > Fear equilibrium variance

---

🧬 Public Audit Transparency

Audit logs (GPT Store deployment target) are written to: /logs/reflexion_audit_YYYYMMDD.json

These logs include CHI scores, tier distributions, and anomaly flags.

Each release includes an ATLAS signature marker.
Signature metadata is regenerated during integrity sweeps.

---

✅ Release Checklist: Integrity + Signatures

If any signed or integrity-audited file is edited:

• Rerun integrity audit workflow
• Regenerate signature metadata
• Do not copy prior signatures forward

Signed/audited artifacts include:

• manifest.json
• BOOTLOADER.md
• bayesian_framework.json
• conversation_starters.json
• stress_test_framework.json
• tags_index.json
• kb_updates_cumulative.json
• knowledge_base_merged_v2.json
• sources.json

A version bump is incomplete until audit metadata and signatures are refreshed for all edited signed artifacts. A version bump is also incomplete until manifest-listed live docs/ surfaces and release-reviewed governance surfaces (BOOTLOADER.md, PROMPTS/guardian_prompt.md) have been explicitly reviewed for package alignment. Run node scripts/validate_kb.js after release-surface edits, and use node scripts/refresh_release_signatures.js --all --write before sealing when signed surfaces changed.

---

🪶 Version Changelog

2.14.0 — NexGen Axiom Guard Integration (2026-04-29)
– Added tools/axiom_guard/ as a bounded governance sidecar.
– Added scripts/axiom_preflight.py release-preflight bridge.
– Added negative-null smoke tests for non-detection claims.
– Registered Axiom Guard in manifest.json.
– Preserved AAIV as T4-only, opt-in, natural-first, and low-prior.
– Breaking changes: none.

2.12.3 — Governance Hardening + Seal Finalization (2026-02-22)
– Canonical surfaces reconciled to sealed state.
– Option C gate + STF infrastructure integrated and hardened.
– Pre-seal citation cleanup completed (free_text_in_citation_fields: 0).

• Composite placeholder sources normalized into quarantine with strict record-shape enforcement (urls non-empty or path).
  – Final deterministic STF full suite logged (total: 22, passed: 11, failed: 0, skipped: 11).

2.12.2 — Truth Surface Reconciliation Patch (2026-02-15)
– Governance reconciliation baseline for 2.12.3 hardening.

2.12.0 (2025-12-05)
– Added AAIV protocol scaffolding (T4-only, opt-in, natural-first).

2.11.2 (2025-12-04)
– Plain Mode documentation
– Audit footer toggle
– Ephemeris helper standardization
– Rumor-handling clarifications

2.11.1 (2025-12-02)
– CHI integration
– Reflexion validation loop
– Continuity tag expansion
– Public audit log policy

2.10.x — Legacy Phase-9 (BAAM core + tier weighting)

---

🧭 GPT Store Integration Statement

This GPT operates under BordneAI’s Love > Fear Constant and Reflexion-based self-audit.
Responses are tier-labeled, provenance-tracked, and continuity-validated.

---

⚖️ License & Attribution

© 2026 BordneAI – 3i/ATLAS Gateway Guide v2.14.0
Released under CC BY-NC-SA 4.0
Integrity Score 9.7
Signature Status: signature_validated