A daily crypto, finance, and AI vocabulary grouping puzzle covering backing, market structure, agent workflows, and retrieval systems.
Stablecoin Reserves
These words describe how a dollar-linked token stays redeemable and on peg.
- Reserve Asset: A reserve asset is held to back a stablecoin or other financial obligation.
- Proof of Reserves: Proof of reserves is a method for demonstrating that an issuer or custodian holds assets that back liabilities, often using attestations and sometimes cryptographic proofs.
- Redemption Window: A redemption window is the period or process used to exchange a token back for its backing asset.
- Depeg Risk: Depeg risk is the chance that a stablecoin moves away from its intended reference value.
Market Depth and Fills
These words describe how orders interact with market depth and trade execution.
- Bid-Ask Spread: The bid-ask spread is the gap between the highest buy price and the lowest sell price in a market.
- Market Depth: Market depth shows how much buy and sell interest exists near the current price.
- Limit Order: A limit order only executes at a chosen price or better.
- Slippage: Slippage is the gap between an expected fill price and the price a trade actually gets.
Agent Loop Control
These words cover the loop an AI agent uses to plan, call tools, store context, and stay constrained.
- Planner: A planner is the part of an AI system that breaks a goal into steps.
- Tool Call: A tool call is a request from a model to use an external function or API.
- Memory: Memory is stored context an agent can reuse while it keeps working.
- Guardrail: A guardrail is a rule or filter that keeps a system within safe bounds.
Retrieval Context Ops
These words describe how applications store knowledge and manage context efficiently.
EmbeddingVector StoreContext WindowFeature Store
- Embedding: An embedding is a numeric representation that lets a model compare meaning across items.
- Vector Store: A vector store is a database for saving and searching embeddings.
- Context Window: A context window is the amount of input and generated text a model can consider at one time when producing a response.
- Feature Store: A feature store manages reusable input variables for machine learning models.