Quick Start •
Why SCBE? •
Use Cases •
Pricing •
Docs

---

The Problem

AI agents are getting autonomous. They can browse the web, execute code, and make decisions. But how do you stop one before it does something harmful?

Traditional approaches fail:
- Rule-based systems → Agents find loopholes
- Linear thresholds → Easy to game with small steps
- Pattern matching → Can't detect novel attacks

The Solution

SCBE-AETHERMOORE uses hyperbolic geometry to create a security boundary where:

> The further an agent drifts from safe behavior, the exponentially harder it becomes to continue.

Think of it like a gravity well - small deviations are cheap, but adversarial behavior gets pulled back or stopped entirely.

``Safe Center (low cost) ───────► Boundary (infinite cost) ●━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━● ALLOW QUARANTINE DENY`

`Results`

| Metric | SCBE | Traditional ML | Rule-Based | |--------|------|----------------|------------| | Detection Rate | 95.3% | 89.6% | 56.6% | | False Positives | 2.1% | 8.3% | 15.2% | | Quantum-Safe | Yes | No | No | | Explainable | Yes | Limited | Yes |

`Quick Start`

`$3`

`bash npm i scbe-aethermoore`

`$3`

`bash pip install scbe-aethermoore`

`python from scbe_aethermoore import validate_action

`Check if an agent action is safe`


result = validate_action(
    agent_id="agent-001",
    action="execute_code",
    context={"trust_score": 0.7, "sensitivity": 0.8}
)

print(result.decision) # ALLOW, QUARANTINE, or DENY print(result.risk_score) # 0.0 to 1.0 print(result.explanation) # Human-readable reason`

`CLI Quick Start (Six Tongues + GeoSeal)`

`bash

`From npm (requires Python 3.x on PATH)`


npx scbe encode --tongue KO --text "hello" > spell.txt
npx scbe decode --tongue KO --as-text --in spell.txt
Or run directly with Python

Encode/Decode (Sacred Tongues)

python scbe-cli.py encode --tongue KO --text "hello" > spell.txt
python scbe-cli.py decode --tongue KO --as-text --in spell.txt
Cross-translate KO → AV

python scbe-cli.py xlate --src KO --dst AV --prefix --in spell.txt
Blend/Unblend pattern (KO:2,AV:1,DR:1)

echo -n "secret" | python scbe-cli.py blend --pattern KO:2,AV:1,DR:1 > blend.txt
python scbe-cli.py unblend --as-text --in blend.txt
GeoSeal envelope (context-aware sealing)

echo -n "payload" | python scbe-cli.py geoseal-encrypt --ss1 > envelope.json
python scbe-cli.py geoseal-decrypt --as-text --in envelope.json
Self-test

python scbe-cli.py selftest


Capabilities
- 14-layer hyperbolic pipeline with risk-gated decisions (ALLOW / QUARANTINE / DENY)
- Sacred Tongues tokenizer with 6x256 bijective encodings
- GeoSeal envelopes and RWP protocol for tamper-evident sealing
- Fleet governance with weighted consensus and rogue detection
- Post-quantum hooks (ML-KEM / ML-DSA integration points)
- Observability via structured telemetry and benchmarks

Full list in docs/CAPABILITIES.md.

`Why SCBE?`

`$3`


- 14-layer pipeline with 12 verified axioms
- Hyperbolic geometry creates exponential cost barriers
- No "boundary creep" - attacks can't inch their way in
$3

- Kyber768 key exchange (NIST approved)
- ML-DSA-65 signatures (Dilithium)
- Ready for the quantum computing era
$3

- Monitor hundreds of AI agents simultaneously
- Automatic quarantine of suspicious behavior
- Full audit trail with cryptographic proofs
$3

- Python SDK, TypeScript SDK
- REST API for any language
- AWS Lambda ready
Use Cases
$3

Control AI agents that browse the web, execute code, or interact with APIs. Stop them before they go off-script.
$3

Manage fleets of specialized AI agents (researchers, coders, reviewers) with trust-based permissions.
$3

Fraud detection, trading bots, risk assessment - with audit trails that satisfy regulators.
$3

Patient data access control with HIPAA-grade security and explainable decisions.
$3

Classified system access control with post-quantum cryptography.
API Example

`bash

`Start the API server`


SCBE_API_KEY=demo-key python -m uvicorn api.main:app --port 8000
Validate an action

curl -X POST http://localhost:8000/v1/validate \
  -H "Content-Type: application/json" \
  -d '{
    "agent_id": "agent-001",
    "action": "read_file",
    "target": "/etc/passwd",
    "trust_score": 0.5
  }'

Response:`json { "decision": "DENY", "risk_score": 0.94, "explanation": "High-sensitivity target with insufficient trust", "layer_scores": [0.5, 0.6, 0.7, 0.8, 0.9, 0.94], "recommendation": "Require human approval for system file access" }`

`Architecture`

`INPUT → [14-Layer Hyperbolic Pipeline] → DECISION

Layer 1-4: Context embedding into Poincaré ball Layer 5-7: Hyperbolic distance + Möbius transforms Layer 8-10: Multi-well realms + spectral coherence Layer 11-12: Temporal distance + Harmonic wall (exponential cost) Layer 13: Risk decision gate → ALLOW / QUARANTINE / DENY Layer 14: Audit telemetry`

`Pricing`

| Tier | Price | Requests/mo | Support | |------|-------|-------------|---------| | Open Source | Free | Unlimited (self-hosted) | Community | | Cloud Starter | $99/mo | 100,000 | Email | | Cloud Pro | $499/mo | 1,000,000 | Priority | | Enterprise | Contact | Unlimited | Dedicated |

`Documentation`

- Getting Started Guide - Technical Reference - API Reference - Architecture Deep Dive - Capabilities

`Contributing`

We welcome contributions! See CONTRIBUTING.md for guidelines.

`bash git clone https://github.com/issdandavis/SCBE-AETHERMOORE.git cd SCBE-AETHERMOORE npm install && pip install -r requirements.txt npm test && pytest``

License

Dual License - See LICENSE for full terms.

- Free for research, education, and personal use
- Commercial use requires a paid license

Patent Pending - USPTO Application #63/961,403

---

Feedback, Support, and Links

| Resource | Link |
|----------|------|
| npm | https://www.npmjs.com/package/scbe-aethermoore |
| Source | https://github.com/issdandavis/SCBE-AETHERMOORE |
| Pitch Deck | https://scbe-aethermoore-x8vu3ly.gamma.site/ |
| Demo | https://scbe-aethermoore-issdandavis.streamlit.app |
| Issues | https://github.com/issdandavis/SCBE-AETHERMOORE/issues |
| Email | issdandavis@gmail.com |
| Sponsor | https://github.com/sponsors/issdandavis |
| CLI Tool | https://aethermoorgames.gumroad.com/l/dqncbs |
| Subscribe | https://aethermoorgames.gumroad.com/subscribe |

---

SCBE-AETHERMOORE

Built for the age of autonomous AI.

Star this repo •
Report Issue •
Contact

---

Quick Start •
Why SCBE? •
Use Cases •
Pricing •
Docs

---

The Problem

AI agents are getting autonomous. They can browse the web, execute code, and make decisions. But how do you stop one before it does something harmful?

Traditional approaches fail:
- Rule-based systems → Agents find loopholes
- Linear thresholds → Easy to game with small steps
- Pattern matching → Can't detect novel attacks

The Solution

SCBE-AETHERMOORE uses hyperbolic geometry to create a security boundary where:

> The further an agent drifts from safe behavior, the exponentially harder it becomes to continue.

Think of it like a gravity well - small deviations are cheap, but adversarial behavior gets pulled back or stopped entirely.

`Results`

`Quick Start`

`$3`

`bash npm i scbe-aethermoore`

`$3`

`bash pip install scbe-aethermoore`

`python from scbe_aethermoore import validate_action

`Check if an agent action is safe`


result = validate_action(
    agent_id="agent-001",
    action="execute_code",
    context={"trust_score": 0.7, "sensitivity": 0.8}
)

print(result.decision) # ALLOW, QUARANTINE, or DENY print(result.risk_score) # 0.0 to 1.0 print(result.explanation) # Human-readable reason`

`CLI Quick Start (Six Tongues + GeoSeal)`

`bash

`From npm (requires Python 3.x on PATH)`


npx scbe encode --tongue KO --text "hello" > spell.txt
npx scbe decode --tongue KO --as-text --in spell.txt
Or run directly with Python

Encode/Decode (Sacred Tongues)

python scbe-cli.py encode --tongue KO --text "hello" > spell.txt
python scbe-cli.py decode --tongue KO --as-text --in spell.txt
Cross-translate KO → AV

python scbe-cli.py xlate --src KO --dst AV --prefix --in spell.txt
Blend/Unblend pattern (KO:2,AV:1,DR:1)

echo -n "secret" | python scbe-cli.py blend --pattern KO:2,AV:1,DR:1 > blend.txt
python scbe-cli.py unblend --as-text --in blend.txt
GeoSeal envelope (context-aware sealing)

echo -n "payload" | python scbe-cli.py geoseal-encrypt --ss1 > envelope.json
python scbe-cli.py geoseal-decrypt --as-text --in envelope.json
Self-test

python scbe-cli.py selftest


Capabilities
- 14-layer hyperbolic pipeline with risk-gated decisions (ALLOW / QUARANTINE / DENY)
- Sacred Tongues tokenizer with 6x256 bijective encodings
- GeoSeal envelopes and RWP protocol for tamper-evident sealing
- Fleet governance with weighted consensus and rogue detection
- Post-quantum hooks (ML-KEM / ML-DSA integration points)
- Observability via structured telemetry and benchmarks

Full list in docs/CAPABILITIES.md.

`Why SCBE?`

`$3`


- 14-layer pipeline with 12 verified axioms
- Hyperbolic geometry creates exponential cost barriers
- No "boundary creep" - attacks can't inch their way in
$3

- Kyber768 key exchange (NIST approved)
- ML-DSA-65 signatures (Dilithium)
- Ready for the quantum computing era
$3

- Monitor hundreds of AI agents simultaneously
- Automatic quarantine of suspicious behavior
- Full audit trail with cryptographic proofs
$3

- Python SDK, TypeScript SDK
- REST API for any language
- AWS Lambda ready
Use Cases
$3

Control AI agents that browse the web, execute code, or interact with APIs. Stop them before they go off-script.
$3

Manage fleets of specialized AI agents (researchers, coders, reviewers) with trust-based permissions.
$3

Fraud detection, trading bots, risk assessment - with audit trails that satisfy regulators.
$3

Patient data access control with HIPAA-grade security and explainable decisions.
$3

Classified system access control with post-quantum cryptography.
API Example

`bash

`Start the API server`


SCBE_API_KEY=demo-key python -m uvicorn api.main:app --port 8000
Validate an action

curl -X POST http://localhost:8000/v1/validate \
  -H "Content-Type: application/json" \
  -d '{
    "agent_id": "agent-001",
    "action": "read_file",
    "target": "/etc/passwd",
    "trust_score": 0.5
  }'

`Architecture`

`INPUT → [14-Layer Hyperbolic Pipeline] → DECISION

`Pricing`

`Documentation`

- Getting Started Guide - Technical Reference - API Reference - Architecture Deep Dive - Capabilities

`Contributing`

We welcome contributions! See CONTRIBUTING.md for guidelines.

`bash git clone https://github.com/issdandavis/SCBE-AETHERMOORE.git cd SCBE-AETHERMOORE npm install && pip install -r requirements.txt npm test && pytest``

License

Dual License - See LICENSE for full terms.

- Free for research, education, and personal use
- Commercial use requires a paid license

Patent Pending - USPTO Application #63/961,403

---

Feedback, Support, and Links

---

SCBE-AETHERMOORE

Built for the age of autonomous AI.

Star this repo •
Report Issue •
Contact

---