All 53 articles, sorted alphabetically
Agent Evaluation at Scale
From spot-checks to continuous monitoring.
Read article →Agent-to-Human Handoff Patterns
When and how to escalate well.
Read article →Agent Memory Layers Compared
Working, episodic, semantic — what to persist where.
Read article →Agent Memory Systems
Short-term context vs long-term vector store vs structured memory.
Read article →Multi-agent debate architecture
Deep-dive on multi-agent debate: two proposers with distinct personas, critic attacks, rebuttal, judge picks winner, cost.
Read article →Agent checkpointing
Deep-dive on checkpointing long-running agents: write-ahead event logs at step boundaries, snapshot-plus-tail resume, idempotency keys and reconciliat…
Read article →Context compaction for long-running agents
Deep-dive on context compaction: keeping a long agent session inside the context window by summarizing settled history rather than truncating it. Cove…
Read article →Context engineering architecture
Deep-dive on context engineering for agents: per-turn token budgets, history windowing and commitment-preserving compaction, tool-result hygiene and a…
Read article →Agent cost optimization -- controlling the token bill
Deep-dive on agent cost optimization: why agents are token-hungry (loops, growing context), the cost drivers, model routing, context management, promp…
Read article →Agent Evaluation
How to evaluate agents: task success, tool usage correctness, trajectory analysis, efficiency.
Read article →Agent Evaluation Architecture in Depth
A 2500-word walkthrough of agent evaluation: task suite, outcome + trajectory, LLM judge, sandbox, adversarial, regression gates, continuous online.
Read article →Agent evaluator architecture
Deep-dive on agent evaluators: task suite, rubric, judge model, trace scoring, regression gate, calibration, multi-judge.
Read article →Agent Frameworks
The major agent frameworks and their design philosophies.
Read article →Agent guardrails
Deep-dive on agent guardrails: input/output/action/budget control points, deterministic vs LLM-based checks, fail modes (block/warn/escalate), defense…
Read article →Human-in-the-loop -- humans and agents collaborating
Deep-dive on human-in-the-loop agents: approval gates, escalation, review/correction, and confidence-based involvement, interaction patterns (sync/asy…
Read article →Agent memory architecture
Deep-dive on agent memory: working, episodic, semantic stores, vector retrieval, consolidation, forgetting, privacy, and user controls.
Read article →Memory Pattern
How agents use short-term (conversation) and long-term (cross-session) memory to be more useful.
Read article →Multi-Agent Pattern
How multi-agent systems decompose work across specialist agents with a coordinator.
Read article →Multi-Agent Orchestration Architecture in Depth
A 2500-word walkthrough of multi-agent orchestration: supervisor, workers, shared memory, message bus, budgets, and human-in-the-loop.
Read article →Agent orchestrator architecture
Deep-dive on agent orchestrator: coordinator, task graph DAG, specialized agents, state store, retries, compensation, budgets.
Read article →Agent output verification architecture
Deep-dive on verifying agent outputs before acting: structure, grounding, and policy layers; deterministic rules vs LLM-as-judge; the pass/repair/reje…
Read article →AI Agents Overview
What AI agents are, how they differ from chatbots, and the loop that defines agentic behavior.
Read article →Agent planner architecture
Deep-dive on agent planner architecture: the reflection loop with state store, memory, guardrails, and observability.
Read article →Agent planner v2 architecture
Deep-dive on agent planner v2: hierarchical goals, actions, state store, adaptive re-plan, budgets, eval hooks, human handoff.
Read article →Planning Pattern
How planning agents decompose complex tasks into subtasks before execution, improving reliability.
Read article →Agent Planning Architecture in Depth
A 2500-word walkthrough of agent planning: task prompt, strategy (ToT/ReAct/Plan-Exec), reflection, replanning, sub-agents, constraint enforcement, st…
Read article →Agents in Production
The practical concerns of running agents in production: monitoring, safety, cost control, incident response.
Read article →ReAct -- reasoning and acting interleaved
Deep-dive on the ReAct agent pattern: interleaving reasoning and acting (thought/action/observation), the loop, grounding in real results, adaptivity …
Read article →ReAct Pattern
How the ReAct pattern (Reasoning + Acting) structures agent loops with explicit thought traces.
Read article →Reflection Pattern
How reflection loops improve agent output by having a critic evaluate and revise.
Read article →Agent reflection architecture
Deep-dive on agent reflection: self-critique, verifier tools, revise loop, bounds, convergence, diverse critics, ranking, evaluation.
Read article →Agent request router architecture
Deep-dive on request routing for agent systems: a cheap classifier that scores intent and difficulty, a routing policy that dispatches to a cheap mode…
Read article →Agent tool sandboxing -- containing what agents can do
Deep-dive on agent tool sandboxing: the untrusted-agent threat (LLM-driven, prompt-injectable), sandboxing tool execution, isolation levels (process/c…
Read article →Agent semantic cache architecture
Deep-dive on semantic caching for LLM agents: embedding-based nearest-neighbor lookup, the similarity threshold as the precision/hit-rate dial, scope …
Read article →Speculative tool execution for agents
Deep-dive on speculative tool execution in LLM agents: predicting the next tool call, running read-only tools eagerly in a sandbox, committing on a ma…
Read article →Agent state machine architecture
Deep-dive on modeling an LLM agent as an explicit finite state machine: named states (intake, plan, act, observe, verify, respond, escalate, error) wi…
Read article →Agent task decomposition
Deep-dive on agent task decomposition: a decomposer that plans subtasks and dependencies, a scheduler that releases the ready set and exploits paralle…
Read article →Agent tool-call recovery architecture
Deep-dive on making agent tool calls robust: an error taxonomy (transient / invalid-input / permanent), idempotency keys for safe replay, exponential …
Read article →Agent tool schema versioning architecture
Deep-dive on versioning agent tool schemas: why a schema is a prompt not an API contract, immutable versioned registries, compatibility classification…
Read article →Tool selection architecture
Deep-dive on dynamic tool selection (tool retrieval) for agents whose catalog has grown to hundreds or thousands of tools, where listing them all in t…
Read article →Agent tool use architecture
Deep-dive on agent tool use: parallel dispatch, streaming, aggregator, error contracts, retries, timeouts, fallback, observability.
Read article →Tools Pattern
How the tools pattern lets agents call functions to affect the world, and how to design good tools.
Read article →LangGraph for Agent Orchestration
State machines for multi-step LLM workflows with checkpointing.
Read article →Multi-Agent Collaboration & Negotiation: Orchestrating Collective Intelligence
Multi-Agent Collaboration & Negotiation: Orchestrating Collective Intelligence
Read article →Multi-Agent Communication Protocols: The Language of Collaboration
Multi-Agent Communication Protocols: The Language of Collaboration
Read article →Multi-Agent Systems: What Happens When a 'Developer' LLM Talks to a 'Reviewer' LLM?
Read article →ReAct Pattern Explained
Reason + Act loops for LLM agents and where they break.
Read article →ReAct Pattern Revisited
Reasoning + acting in 2026 — what survived.
Read article →The Rise of 'Thinking' Models: How Chain-of-Thought (CoT) Is Turning LLMs into Logic Engines
Read article →Tool Calling Best Practices
Designing tool descriptions retry policies and timeout handling.
Read article →Tool Use Failure Modes
Six ways agents call tools wrong.
Read article →