How Σternal Thinks About Markets
Σternal approaches markets as evolving environments rather than sequences of isolated events.
Price, policy, liquidity, and behavior interact over time. Meaning emerges when current conditions are placed inside a longer historical record of similar environments.
Σternal exists to provide that placement.
It helps users recognize the type of market they are operating in, how comparable environments unfolded in the past, and where the present appears aligned or structurally different.
The system is designed to support judgment, not replace it.
Getting Started
Using Σternal begins with a simple action.
Select an asset. Select a timeframe.
From there, Σternal evaluates the surrounding market environment and compares it to historical regimes that share similar structural characteristics.
The output is a structured view of context that can be revisited as conditions evolve.
System Overview
Σternal decomposes markets into a small number of persistent dimensions that have historically shaped outcomes across cycles.
For each decode, the system:
- •Assesses the current environment across core dimensions
- •Identifies historical regimes with comparable structure
- •Surfaces relevant past periods and their dynamics
- •Highlights alignment, divergence, and transition risk
This creates a stable reference point for reasoning under uncertainty.
Methodology
The system is built on three guiding principles.
Context
Markets are interpreted relative to prior environments rather than in isolation. Conditions are evaluated as part of a broader system.
Memory
Historical outcomes are treated as evidence. Both continuity and failure of analogies are preserved and surfaced.
Orientation
Outputs are designed to improve situational awareness and support deliberate decision making.
Regime Analysis
A regime is a period defined by a relatively stable configuration of structural forces.
Σternal evaluates regimes across five primary dimensions:
Each dimension is assessed independently and combined to form an environment profile.
Similarity Scoring
Current environments are compared to historical regimes using a weighted, multi-dimensional similarity process.
Similarity expresses structural resemblance, not outcome expectation. Comparable conditions may still produce different results depending on sequencing, policy response, and external shocks.
Scores are intended to guide interpretation rather than imply repetition.
Uncertainty
Market environments are often transitional or internally inconsistent.
Σternal reflects this by attaching confidence levels to its assessments. Lower confidence indicates overlapping regimes, rapid change, or limited historical precedent.
Uncertainty is treated as information rather than error.
Decoder Outputs
Each decode produces a consistent set of outputs.
Environment Snapshot
A concise assessment of current conditions across each dimension.
Regime Distribution
A weighted set of historical regime matches.
Historical Parallels
Specific past periods with contextual explanation.
Similarities and Differences
Where comparisons align and where they diverge.
Scenario Considerations
Potential transition paths and observable inflection points informed by historical behavior.
Scope
Σternal is designed to strengthen understanding of market environments over time.
Historical comparison provides structure and perspective, but markets remain adaptive systems shaped by novelty and change. Outputs should be used as a reference layer that compounds insight rather than compresses uncertainty into answers.
API for Developers
All Tools-menu capabilities are available through the API.
Overview
The `/developers` entry point makes Σternal workflows directly usable from an application, backend service, script, or agent.
It exposes the same working surfaces available in Tools: decode, news, compare, narrative, monitoring, scenarios, research, laboratory, and reference resources such as the regime library.
On LLM-powered endpoints, you can force a `model_name` per request. If you do not, the API uses the default model attached to the account or API key.
Authentication
Create API keys from `/account/api-keys`. Once generated, they can call every `/developers` route allowed by your plan.
Two authentication formats are accepted. `X-API-Key` works well for straightforward integrations. `Authorization: Bearer` is useful if your infrastructure already standardizes on that format.
curl -X POST https://eternal-agents.com/developers/api/v1/decode \
-H "X-API-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"asset": "SPY", "timeframe": "6M", "model_name": "gpt-5-mini"}'A good first call for verifying the key, response format, and model selection.
Endpoints
The public surface is organized into coherent families: decode and technical, news and reports, compare, narrative, monitoring, scenarios, research, personal context, regimes, laboratory, and price data.
/developers/api/v1/decode/developers/api/v1/chat/developers/api/v1/decode/sessions/developers/api/v1/decode/session/{session_id}/developers/news/sources/developers/news-reports/developers/news-reports/jobs/developers/api/v1/decode/jobs/developers/compare/*/developers/narrative/*/developers/monitoring/*/developers/scenarios/*/developers/research/*/developers/regimes/*/developers/personal/*/developers/api/v1/lab/*/developers/api/v1/prices/{ticker}Rate Limits
Calls are governed by your subscription and credit budget. Cost depends on the effective LLM workload and the model you select.
https://eternal-agents.com/developers/openapi.json