AI Agents
KEK uses a system of specialized AI agents to analyze markets, generate strategy hypotheses, and support structured validation across the strategy lifecycle.
Each agent performs a narrowly defined role. No agent executes trades, controls capital, or bypasses validation.
Agents exist to produce intelligence — not action.
What this page covers
- The purpose of AI agents in KEK
- Core agent roles and responsibilities
- How agents coordinate through MCP
- How agent outputs flow into validation layers
- Execution and custody safety boundaries
Purpose of AI agents in KEK
KEK's agents are designed to accelerate research and strategy development while preserving a validation-first system:
- Agents generate structured insights and testable strategy candidates
- Strategies must pass backtesting and paper trading before any execution is eligible
- Execution, when chosen, remains explicitly user-authorized and non-custodial
Backtesting simulates a strategy using historical data to analyze results and risk before risking capital.
Paper trading is simulated trading that allows observing behavior without real money at risk.
Design philosophy
KEK's agent system is built around three principles:
1) Specialization over general intelligence
Each agent is optimized for a specific domain (regime detection, narrative context, asset relevance, strategy synthesis).
2) Coordination over autonomy
Agents coordinate through shared, structured interfaces rather than acting independently or "free-running" in production.
3) Validation over execution
Agent outputs are never executable by default. They become eligible for execution only through the validation pipeline.
Agent architecture overview
Each agent focuses on a single domain of analysis and produces structured outputs that flow downstream.
Agents do not:
- Trade
- Allocate capital
- Bypass validation
- Act independently in production
Their outputs are inputs to the broader strategy lifecycle.
Core agents
Market Regime Agent
Purpose: Identify the prevailing market regime and detect transitions.
Inputs
- Market-wide price behavior
- Volatility and dispersion metrics
- Structural and trend/range indicators
Outputs
- Regime classification (e.g., trending, ranging, volatile, transitional)
- Confidence scores and regime-shift alerts
Why it matters
This agent informs how strategies should behave under current conditions — not which assets to trade or when to execute.
Narrative Agent
Purpose: Identify macro, sector, and thematic narratives that influence asset behavior.
Inputs
- Cross-asset structure and correlation signals
- Liquidity and sentiment proxies
- Contextual drivers (risk-on / risk-off behavior, thematic rotations)
Outputs
- Narrative classifications and active themes
- Theme relevance scoring and persistence signals
Why it matters
Narratives guide context and relevance weighting — they do not generate orders or execution decisions.
Asset Relevance Agent
Purpose: Ranking assets most relevant to the current regime and narrative context.
Inputs
- Regime classification and confidence
- Narrative theme relevance
- Asset-level metrics (trend strength, volatility, liquidity, behavior fit)
Outputs
- Ranked asset relevance scores
- Context-aware filtering signals
Why it matters
This agent reduces noise and focuses strategy generation on assets most likely to express the targeted edge.
Strategy Agent
Purpose: Synthesize agent intelligence into structured strategy hypotheses.
Inputs
- Regime context
- Narrative signals
- Asset relevance scores
- User-defined constraints and objectives
Outputs
- Machine-readable strategy specifications
- Parameterized rules, constraints, and assumptions
- Variant templates for validation
Important
These strategies are candidates, not approved deployments. They must pass validation before any execution is possible.
Multi-agent coordination (MCP)
Agents coordinate through the MCP (Model Context Protocol), which provides a structured system for connecting models to tools, data sources, and workflows via standardized interfaces.
MCP enables:
- Shared context across agents and sessions
- Structured message passing between agent roles
- Tool coordination (data retrieval, analysis, validation workflows)
- Versioned outputs and reproducible agent artifacts
This coordination model supports coherent multi-agent behavior without relying on hidden, mutable, or implicit state.
How agent outputs are used
Agent outputs flow into KEK's validation-first pipeline:
Agent Intelligence → Strategy Generation → Backtesting & Optimization → Paper Trading → Monitoring & Meta-Learning → Optional Execution
At no point do agents bypass validation layers.
- Backtesting evaluates strategy behavior under historical conditions.
- Paper trading evaluates behavior under live market conditions without real capital risk.
Boundaries & safety
KEK agents are constrained by design:
- Do not access or custody user funds
- Do not store or handle private keys
- Do not self-deploy strategies
- Do not operate without downstream validation gates
All agent outputs are treated as inputs to structured validation, monitoring, and (optionally) user-authorized execution.
Why this matters
This architecture exists to:
- Prevent black-box behavior by keeping outputs structured and reviewable
- Enable explainable strategy development through specialization
- Reduce systemic risk through validation gating
- Maintain long-term trust via clear responsibility boundaries
KEK treats intelligence as input — not authority.