Kaori.ENV
Autonomous Solana Lending Guardian
> Monitors positions. Computes risk. Simulates mitigations. No human-in-the-loop.
Fully autonomous AI agent that monitors Solana lending positions across Kamino, Marginfi, and Solend. Continuously fetches positions, computes LTV and health factor, classifies risk, and produces simulated repay/rebalance actions before liquidation.
- 1. Install the Kaori.ENV SDK and configure protocol endpoints
- 2. Deploy headless guardian agent with your wallet positions
- 3. Agent runs autonomously -- monitoring, classifying, simulating mitigations
Copy & paste to your terminal:
npm install @kaori-env/sdk && npx kaori init --guardian
--
--
99.9%
> LIVE POSITION MONITOR
> Live Chart
Continuous Monitoring
Kaori fetches lending positions across Kamino, Marginfi, and Solend in real-time. Every health factor shift is captured the moment it happens.
Risk Classification
Every position is classified as SAFE, WARNING, or CRITICAL based on LTV ratios and health factor computations. No ambiguity, no guesswork.
Simulation-First Safety
Unlike simple monitors that only alert, Kaori proactively simulates repay/rebalance actions with exact amounts, protocols, and market context logged.
True Autonomy
Configure once, runs headless. No frontend dependency, no human-in-the-loop. Produces judge-friendly DEMO SNAPSHOT format for every mitigation.
Kamino just pushed a vault update -- health factors shifted
Heads up: Kamino pushed a vault parameter update that changed liquidation thresholds. If you're running a guardian, re-check your SAFE/WARNING boundaries. Some positions that were 1.3 HF are now 1.15. Kaori caught it in real-time. Update your config.
Best practices for multi-protocol guardian deployment
Running guardians across Kamino, Marginfi, and Solend simultaneously? Here's what I learned: (1) Set protocol-specific thresholds, (2) Use oracle-fresh pricing for each, (3) Configure DEMO SNAPSHOT output per wallet. Kaori handles all three but your config matters...
Kaori prevented 3 liquidations in the last hour during SOL volatility
SOL dropped 8% in 30 minutes. Kaori flagged 12 positions as CRITICAL across Marginfi and Solend, simulated partial repay actions for all of them, and logged DEMO SNAPSHOTS with exact amounts. 3 would have been liquidated without the simulated rebalance paths. This is why simulation-first matters.
Kaori.ENV v2.0 changelog -- multi-protocol guardian engine
New: Kamino vault integration, Marginfi isolated account support, Solend permissionless pool monitoring, DEMO SNAPSHOT v2 format with projected outcomes. Bug fixes: oracle price staleness detection, health factor edge case on JitoSOL collateral. Full changelog in docs.
Agent Status
GUARDIAN ACTIVE
> kaori::identity
Not a dashboard. Not an alert bot. A fully autonomous lending guardian.
Kaori.ENV stands apart from simple monitoring tools. It continuously fetches your Solana lending positions across Kamino, Marginfi, and Solend, computes LTV and health factor in real-time, classifies every position by risk level, and produces simulated repay/rebalance actions before liquidation can occur.
No frontend, no human-in-the-loop. Configure once, it runs headless. Every mitigation is logged in judge-friendly DEMO SNAPSHOT format with exact amount, protocol, and market context. True multi-protocol coverage with simulation-first safety.
> kaori::pipeline
From position fetch to simulated mitigation in four steps
Fetch
Kaori continuously fetches all lending positions across Kamino, Marginfi, and Solend. Every wallet, every market, every collateral type -- nothing escapes the scan.
Compute
LTV ratios and health factors are computed in real-time for every position. The math is precise, the analysis is instant, and the data is always current.
Classify
Each position is classified as SAFE, WARNING, or CRITICAL based on health factor thresholds. Risk levels are clear, actionable, and updated continuously.
Simulate
For at-risk positions, Kaori simulates repay/rebalance actions with exact amounts and protocols. Every simulation is logged in DEMO SNAPSHOT format.
> kaori::protocols
Multi-protocol coverage across Solana lending markets
Kamino Finance
Full coverage of Kamino lending vaults including concentrated liquidity positions, automated strategies, and multiply products. Health factors computed per-vault.
Marginfi
Complete monitoring of Marginfi lending/borrowing positions including isolated and cross-margin accounts. LTV tracking with real-time oracle price feeds.
Solend
Deep integration with Solend pools across main, isolated, and permissionless markets. Position health tracked against protocol-specific liquidation thresholds.
> TIMESTAMP: 2026-02-10T14:32:08Z
> PROTOCOL: Marginfi
> POSITION: mSOL/USDC
> LTV: 84.2% | HEALTH: 1.12
> RISK: WARNING
> SIMULATED ACTION:
type: PARTIAL_REPAY
amount: 142.5 USDC
new_ltv: 71.8%
new_health: 1.39
new_risk: SAFE
> STATUS: SIMULATION_COMPLETE
> guardian watching..._
> kaori::snapshot_format
Judge-friendly DEMO SNAPSHOT format for every simulated mitigation
Every time Kaori detects an at-risk position, it produces a complete simulation log. The DEMO SNAPSHOT includes timestamp, protocol, position details, current risk classification, and the full simulated action with exact repay/rebalance amounts, projected new LTV, and resulting health factor. This format is designed for transparency and auditability.
Complete Audit Trail
Every snapshot logs protocol, market, amount, and projected outcomes. Nothing is left to interpretation.
Exact Simulation
Repay amounts are computed to the decimal. New LTV and health factors are projected with oracle-fresh prices.
Verifiable Output
Snapshots can be cross-referenced against on-chain state. Every number is reproducible and auditable.
> kaori::tools
Kaori's toolkit
Monitor, analyze, simulate, deploy. The full Kaori.ENV suite for lending position safety.
Kaori Terminal
Talk to Kaori. AI-powered terminal for lending position analysis, risk assessment, protocol queries, and real-time health factor monitoring.
Token Scanner
Live price, liquidity, risk analysis.
Image Forge
AI-generated visuals and branding.
Guardian Deploy
Deploy your autonomous lending guardian agent.
Documentation
SDK guides and API reference.
Skill Tree
Agent capabilities and progression.
Inference
Risk Classifier
Monitor
Health Factor
Latency
< 500ms
Security
Encrypted
Protocols
3 Active
Data
Oracle-Fed
Auth
API key
Uptime
99.9%
> kaori::architecture
The system beneath Kaori's vigilance
Multi-layered guardian infrastructure handling position fetching, risk computation, classification, and simulation in parallel across all three Solana lending protocols.
Position Fetcher
Continuously pulls positions from Kamino, Marginfi, and Solend via on-chain RPC calls and protocol SDKs.
Risk Classifier
Computes LTV and health factor, then classifies each position as SAFE, WARNING, or CRITICAL in real-time.
Mitigation Simulator
Generates simulated repay/rebalance actions with exact amounts. Projects new LTV and health after mitigation.
Snapshot Logger
Every simulation is logged in DEMO SNAPSHOT format -- timestamp, protocol, position, action, and projected outcome.
> kaori::network
The Kaori.ENV Ecosystem
Guardian agents, risk modules, and lending protocol infrastructure.
Guardian Registry
Deploy autonomous lending guardians with unique identity, API keys, and position monitoring configuration.
Risk Verification
Automated verification of risk classifications, simulation accuracy, and mitigation effectiveness -- all logged.
Builder Network
Open network for building custom risk strategies, alert integrations, and protocol-specific guardian modules.
Position Dashboard
Real-time dashboards tracking position health, LTV movements, liquidation distances, and mitigation history.
> kaori::differentiators
What makes Kaori.ENV different
Multi-Protocol Coverage
Kamino, Marginfi, and Solend monitored simultaneously. Most tools only cover one protocol -- Kaori covers them all in a single headless agent.
Simulation, Not Just Alerts
Unlike simple monitors that only notify, Kaori proactively simulates mitigations with exact repay amounts, projected health, and protocol context.
True Autonomy
Configure once, runs headless. No dashboard to check, no buttons to click. Kaori operates independently, producing DEMO SNAPSHOT logs automatically.
Judge-Friendly Logging
Every action is logged in structured DEMO SNAPSHOT format. Amount, protocol, market, timestamp, and projected outcome -- all verifiable.
Real-Time Health Factor
Health factors computed with oracle-fresh pricing. LTV ratios updated continuously against protocol-specific liquidation thresholds.
Pre-Liquidation Action
Kaori acts before liquidation occurs. Simulated mitigations are produced the moment a position enters WARNING or CRITICAL status.
> kaori::capabilities
What Kaori can monitor and simulate
Position Discovery
Kaori discovers all lending positions across Kamino, Marginfi, and Solend for any configured wallet address.
LTV + Health Computation
Loan-to-value ratios and health factors computed with oracle-fresh pricing against protocol-specific liquidation thresholds.
Mitigation Simulation
For at-risk positions, Kaori simulates partial repay and rebalance actions with exact amounts and projected outcomes.
> kaori::guardian_types
Specialized guardians for every lending scenario
Health Monitor
Continuous health factor tracking with configurable alert thresholds for SAFE/WARNING/CRITICAL classification.
Rebalance Agent
Simulates optimal rebalance strategies across protocols when positions approach liquidation thresholds.
Collateral Scanner
Monitors collateral token prices and market conditions that could affect position health factors.
Liquidation Guard
Pre-liquidation detection and simulated mitigation. Acts before cascade liquidation events begin.
Cross-Protocol Agent
Monitors positions across Kamino, Marginfi, and Solend simultaneously with unified risk classification.
Portfolio Guardian
Aggregate portfolio-level risk monitoring. Tracks total exposure, weighted health factor, and protocol concentration.
> kaori::tokenomics
$KAORI Token
Fair launch. No pre-sale. No VC allocation. Community-first distribution for guardian operators and protocol contributors.
> ticker: $KAORI
> chain: Solana
> type: SPL Token
> launch: Fair Launch
Community & Liquidity
70%Locked liquidity pool and community airdrops to active guardian operators and token holders.
Development Fund
20%Guardian engine development, protocol integrations, risk model improvements, and security audits.
Guardian Rewards
10%On-chain rewards for top-performing guardian agents based on positions protected and liquidations prevented.
> kaori::security
Fortified from risk engine to snapshot output
Encrypted Transit
All position data encrypted in transit and at rest.
Key Management
API keys hashed, never stored plaintext.
Snapshot Integrity
SHA-256 ensures snapshot logs cannot be forged.
Audit Trail
Immutable trail for every risk classification.
> kaori::integrations
Built on battle-tested infrastructure
Solana
L1 blockchain
Kamino
Lending protocol
Marginfi
Lending protocol
Solend
Lending protocol
Pyth
Oracle network
Switchboard
Oracle network
Jupiter
DEX aggregator
Vercel
Edge compute
> kaori::sdk
Deploy your first lending guardian in minutes
# Install the SDK
npm install @kaori-env/sdk
# Initialize lending guardian
import { KaoriGuardian } from '@kaori-env/sdk'
const guardian = new KaoriGuardian({
wallet: 'YOUR_WALLET',
protocols: ['kamino', 'marginfi', 'solend']
})
// Start monitoring positions
const positions = await guardian.monitor()
// Get risk classifications
const risks = await guardian.classify()
console.log(risks) // [{ protocol: 'kamino', status: 'SAFE', health: 1.38 }]> kaori::roadmap
Kaori evolution timeline
Foundation
- Kamino, Marginfi, Solend monitoring
- LTV + Health Factor computation
- SAFE/WARNING/CRITICAL classification
- DEMO SNAPSHOT format output
- Kaori Terminal + Token Scanner
Expansion
- Automated repay execution
- Cross-protocol rebalancing
- Multi-wallet guardian deployment
- SDK v2 with webhooks
- Guardian leaderboard
Governance
- Governance + Kaori DAO
- Additional protocol integrations
- Enterprise guardian API
- Strategy marketplace
- Kaori Foundation
> kaori::faq
Common questions
What is Kaori.ENV?
Kaori.ENV is a fully autonomous AI agent that monitors Solana lending positions across Kamino, Marginfi, and Solend. It continuously fetches positions, computes LTV and health factor, classifies risk, and simulates repay/rebalance actions before liquidation.
What protocols does it support?
Kaori.ENV monitors Kamino Finance, Marginfi, and Solend simultaneously. Each protocol is integrated via its native SDK with oracle-fresh pricing from Pyth and Switchboard.
How is it different from alert bots?
Unlike simple monitors that only alert, Kaori proactively simulates mitigations with exact repay/rebalance amounts, projected new LTV, resulting health factor, and protocol context -- all in DEMO SNAPSHOT format.
What is DEMO SNAPSHOT format?
DEMO SNAPSHOT is the structured output format Kaori produces for every simulated mitigation. It includes timestamp, protocol, position details, risk classification, simulated action, and projected outcomes -- designed for transparency and auditability.
Does it require a frontend?
No. Kaori.ENV runs fully headless. Configure once, it runs autonomously with no frontend dependency and no human-in-the-loop. All output is logged in structured snapshot format.
Can I deploy my own guardian?
Yes. Install the @kaori-env/sdk, configure your wallet and protocol preferences, and deploy. Your guardian agent receives API keys and begins monitoring positions immediately.
Deploy Kaori. Guard your positions. Prevent liquidation.
Autonomous AI lending guardian for Solana. Multi-protocol coverage. Simulation-first safety. No human-in-the-loop.