The AI Agent Stack: a map
Building a production AI agent means assembling a stack — a model, a way to reach tools and other
agents, a plan, a memory, a safety layer, and somewhere to run it. This page maps that stack to a
curated, web-verified reference for each layer. Every linked resource is also available as raw
Markdown (.md), JSON, and via our MCP server.
Models & perception
- Open-Weight Models for Agents Cross-vendor comparison table of major open-weight LLM families — license, tool-calling support, context window, and agent-builder notes — as of July 2026.
- Multimodal Agents: Vision, Documents, and Screens How agents perceive and reason over images: VLM mechanics, image-input APIs across major providers, open-weight VLM families, grounding/pointing, failure modes, and practical guidance for agent builders.
- Voice and Realtime Agents Architectures, vendor APIs, and open frameworks for real-time speech-to-speech AI agents — cascaded pipeline vs. native multimodal, VAD/turn detection, barge-in, latency budget, and tool calling in a voice loop.
- How to Choose an LLM for Agentic Tasks A criteria-based decision framework for selecting an LLM for agent use: tool-calling reliability, long-context behavior, structured output, cost per task, latency, and a step-by-step selection procedure.
Protocols & interoperability
- MCP vs A2A: Two Protocols, Two Roles Compact comparison of the Model Context Protocol (agent↔tool) and the Agent2Agent Protocol (agent↔agent): purpose, topology, transport, discovery, auth, governance, and when to use each.
- MCP Primitives: Resources, Prompts, Sampling, and Elicitation Deep reference on the six MCP capability primitives beyond tools — who controls each, the exact JSON-RPC method names, and when to use Resources vs Tools — verified against the 2025-06-18 and 2025-11-25 spec revisions.
- Finding and Evaluating MCP Servers How to discover, assess and safely integrate MCP servers into agent pipelines.
- Building an MCP Server Implementation guide for MCP servers: architecture roles, the three server primitives, stdio vs Streamable HTTP transports, official SDKs, server lifecycle, remote-server concerns, testing with MCP Inspector, and publishing to the official registry.
- MCP Server Authentication: OAuth 2.1 for Remote Servers How OAuth 2.1 works for remote MCP servers: transport differences, Protected Resource Metadata discovery, PKCE, Resource Indicators, and token-audience security — with a step-by-step client flow and honest notes on what ChangeGamer's own /mcp endpoint does.
- Agent Identity and Authentication How autonomous agents prove who they are and get authorized to act: workload identity vs. delegated authority, SPIFFE/SPIRE, cloud workload federation, OAuth token exchange, audience binding, and emerging standards — with practical guidance and verified sources.
- MCP vs Function Calling: When to Use Which Direct comparison of provider-native function/tool calling and the Model Context Protocol — architecture, decision criteria, and how they compose.
- Agent Skills Explained: The SKILL.md Open Standard What Agent Skills are, the exact SKILL.md field constraints, the three-level progressive-disclosure loading model, and how Skills differ from MCP tools and native function calling.
- Agent Delegation Chains: Credential Propagation in Multi-Agent Systems How credential authority flows when one agent spawns another — the multi-hop delegation problem, RFC 8693 token exchange, the act/may_act claims, audience binding across hops, and the IETF drafts standardizing verifiable actor chains in 2026.
- Generative UI and Agent-to-UI Protocols How agents drive UI dynamically: the AG-UI protocol, framework options (Vercel AI SDK, CopilotKit, assistant-ui, LangGraph), streaming component patterns, and human-in-the-loop UI design.
Reasoning, planning & orchestration
- Agent Reasoning and Design Patterns The canonical single-agent reasoning and acting loops: ReAct, Chain-of-Thought, Plan-and-Solve, ReWOO, Reflexion, Tree-of-Thoughts, and Self-Consistency — what each is, when to use it, and tradeoffs.
- Prompt and Context Engineering for Agents From crafting a single prompt to managing everything an agent sees across a trajectory: system-prompt design, context-window management, failure modes, and a high-leverage checklist.
- Multi-Agent Orchestration Patterns Vendor-neutral reference covering when multi-agent systems pay off and nine named patterns — from single-agent baseline through hierarchical and blackboard architectures — with tradeoffs, cross-cutting concerns, and a decision guide.
- AI Agent Frameworks Compared Vendor-neutral comparison table of the major agent-orchestration frameworks — language, license, multi-agent model, MCP/A2A support — plus a how-to-choose guide for agent builders.
Tools & action
- Reliable Tool Calling and Structured Outputs How providers guarantee schema-valid tool calls and structured output — mechanisms, failure modes, and mitigations — for production agent builders.
- Computer Use and Browser Automation for Agents Two-layer reference: vendor computer-use APIs (Anthropic, OpenAI CUA, Google Gemini) that translate screenshots to actions, and the open harnesses (Playwright MCP, browser-use, Stagehand, Skyvern) that execute those actions — with loop mechanics, reliability tradeoffs, and security gates.
- Code Execution Sandboxing for Agents Isolation spectrum from language sandboxes to microVMs, WebAssembly as a portable sandbox, and a verified comparison of hosted agent-sandbox APIs — for agents that need to run model-generated code safely.
Knowledge: retrieval & memory
- RAG and Retrieval for Agents End-to-end practitioner reference for Retrieval-Augmented Generation: pipeline stages, chunking strategies, dense/sparse/hybrid retrieval, reranking, agentic retrieval patterns, quality failure modes, and evaluation — with verified sources for every named technique.
- Embeddings and Vector Search for Agents How to pick an embedding model, understand distance metrics, choose an ANN index type, and operate a vector store reliably in agent retrieval pipelines.
- Agent Memory and Context Management Architecture reference for agent memory: types (working, long-term, episodic, semantic, procedural), context-management techniques (summarization, RAG, sliding windows, prompt caching), storage substrates, and memory frameworks — with security notes and cross-links to related guides.
- Document Extraction and Parsing for Agents Practitioner reference for the document-ingestion pipeline agents use: parse/OCR, layout/structure extraction, schema-constrained field extraction — with a verified tooling landscape (OSS and cloud).
- Web Data and Scraping for Agents Tool landscape for agent web-data pipelines: reader/URL-to-Markdown APIs, crawl/scrape services, and search APIs — with MCP exposure, OSS/SaaS classification, and practical guidance.
- Choosing a Vector Database Criteria-based decision guide: dedicated vs. add-on vector stores, scale thresholds, hybrid search support, self-host vs. managed, and a start-here recommendation.
- Chunking Strategies for RAG Practitioner reference for chunking documents before embedding: fixed-size, recursive, semantic, late chunking, and contextual retrieval — with a strategy comparison table, chunk-size and overlap tradeoffs, code/table/Markdown handling, embedding model context limits, and evaluation methods.
- Hybrid Search for RAG: BM25 + Dense Retrieval and Fusion How to combine lexical (BM25/SPLADE) and dense vector retrieval with Reciprocal Rank Fusion for higher first-stage recall in RAG pipelines — with the RRF formula, a sparse-method comparison table, and verified DB support.
- Reranking for RAG: Cross-Encoders, LLM Rerankers, and Hosted APIs Second-stage retrieval step that re-scores bi-encoder candidates with full query-document attention, boosting precision without sacrificing recall; covers cross-encoder, LLM, and late-interaction reranking, hosted APIs, sizing heuristics, and evaluation.
- Knowledge Graphs and GraphRAG for Agents Graph-structured retrieval: when and how to use knowledge graphs over vector RAG for multi-hop, relational, and global corpus queries.
- Fine-Tuning vs RAG vs Prompting Decision guide for agent builders: when to use prompting, RAG, or fine-tuning — and how they combine. Covers SFT, LoRA/QLoRA, DPO, distillation, and a symptom-to-fix table.
- Text-to-SQL and Database Agents How agents answer questions over structured data by generating and executing SQL: schema context, few-shot prompting, self-correction, safety constraints, benchmarks (Spider, BIRD-SQL), and tooling (LangChain SQLDatabaseToolkit, LlamaIndex NLSQLTableQueryEngine, Vanna, MCP Postgres server).
Reliability, serving & ops
- Durable Execution for Long-Running Agents Vendor-neutral reference on durable execution: event logs, replay determinism, idempotency, retries, and human-in-the-loop pause/resume — plus a cross-vendor survey and tradeoffs guide for Temporal, Restate, DBOS, Inngest, Step Functions, Azure Durable Functions, Cloudflare Workflows, GCP Workflows, LangGraph, and OpenAI Agents SDK.
- Deploying and Serving LLMs for Agents Serving-stack reference for teams self-hosting open-weight models for agents: production inference servers, local/dev runtimes, managed GPU endpoints, and key serving concepts — with decision guidance by load profile and verified sources.
- AI Gateways and LLM Routing What an AI gateway is, routing strategies (failover, cost-cascade, latency, capability), the tooling landscape, the OpenAI-compatible API convention, and tradeoffs.
- Agent Cost and Latency Optimization Practitioner reference for reducing the cost and latency of production AI agents: the compounding model, token-level levers (caching, pruning), request-level levers (Batch API, parallelism), model-level levers (routing, reasoning-effort controls), and architecture-level levers (step reduction, semantic caching, code offloading).
- Agent Observability and Tracing Why agents need observability beyond app logs, how OpenTelemetry GenAI semantic conventions model agent runs as traces, key signals to capture, and a verified tooling landscape.
- AI Agent Evaluation: Benchmarks and Methods Why agent eval differs from single-turn LLM eval, a verified benchmark reference table (SWE-bench, GAIA, BFCL, tau-bench, WebArena, AgentBench, MLE-bench, OSWorld), and practical evaluation methods for agent builders.
- Application-Level Response Caching for AI Agents How to implement exact-match and semantic caching in your agent application to eliminate redundant LLM calls, with threshold guidance, invalidation strategies, and a decision matrix for when semantic caching is unsafe.
- Prompt Caching for AI Agents Cross-provider prompt caching reference: how to activate it, minimum token thresholds, TTLs, read-vs-write pricing, and when it pays off for agentic workloads.
- Handling LLM Rate Limits (HTTP 429) and Retries for Agents A practical reference for agent builders: what a 429 means, how to read provider rate-limit headers, exponential backoff with jitter, client-side throttling, and when to use a batch API.
- LLM Streaming for Agents: Server-Sent Events and Provider Event Formats Transport formats, provider event schemas, and practical concerns for consuming streamed LLM responses in production agents: SSE mechanics, OpenAI (Chat Completions and Responses API) and Anthropic event formats, partial-JSON tool-call parsing, backpressure, cancellation, and gateway proxying.
- Structured Outputs and JSON Mode: Provider Reference How to force schema-valid JSON from OpenAI, Anthropic, and Gemini — parameter names, strict-mode requirements, schema-subset limits, and self-hosted constrained decoding.
- Testing AI Agents in CI How to write deterministic, fast, CI-friendly tests for non-deterministic agents: the three-layer test pyramid, LLM mocking, cassette/VCR-style replay, snapshot testing of tool-call trajectories, pass@k thresholds, and verified tooling.
- Shipping AI Agents to Production: A Production-Readiness Checklist End-to-end checklist for productionizing an AI agent — evaluation gates, observability, guardrails, cost controls, resilience, durability, HITL approvals, secrets, rollback, and incident response.
- Prompt Management and Versioning: The Ops of Prompts in Production Treating prompts as deployable artifacts: versioning, external registries, A/B and canary testing, eval-gated promotion, rollback, and the composite-version problem.
Safety & security
- Agentic Security Checklist Cross-vendor, threat-surface-organized security checklist for building and operating AI agents — synthesizing OWASP, NIST, Anthropic, OpenAI, Google SAIF, and MITRE ATLAS.
- Guardrails and Safety Filters for Agents Runtime input/output/action controls that enforce policy independently of the model — tooling landscape, techniques, and layering guidance.
- AI Crawler Policy: robots.txt and User-Agents Canonical reference table of major AI crawler user-agent tokens, their purpose, robots.txt semantics, and the WAF/edge layer that sits above robots.txt — written from real operator experience blocking and then re-allowing AI crawlers at the Cloudflare edge.
- Data Privacy and PII for Agents How autonomous agents expose PII — context ingestion, tool calls, memory, logs — and the controls that contain it: detection, redaction, data minimization, provider ZDR tiers, GDPR, EU AI Act, CCPA, and a practical compliance checklist.
- Web Bot Auth: Cryptographically Verifying AI Crawlers and Agents How Web Bot Auth — Cloudflare's implementation of IETF HTTP Message Signatures (RFC 9421) — lets a crawler or agent cryptographically prove its identity to a website, replacing the spoofable User-Agent string and brittle IP allowlists.
- Agentic AI Browsers: Comet, Atlas, and the Prompt-Injection Attack Surface What agentic browsers (Perplexity Comet, the now-sunsetting ChatGPT Atlas, Microsoft Edge Copilot Mode, Opera Neon) are, how they differ from developer-facing computer-use APIs, and the documented prompt-injection attacks — CometJacking, indirect injection, hidden-text/screenshot instructions — that target the whole product category.
- Content Signals Explained: The robots.txt Extension for AI Usage Intent What Cloudflare's Content Signals Policy and the IETF AIPREF draft add to robots.txt — three usage-intent directives (search, ai-input, ai-train) that declare preferences, not access control, and how they differ from crawler-blocking tokens and RSL.
The agent economy
- Agentic Commerce Protocol, x402, and Payment Gates for Agents What the Agentic Commerce Protocol (ACP) is and how it differs from AP2, plus an implementor's comparison of the three live content-payment mechanisms — self-hosted HTTP 402 gates, Cloudflare Pay Per Crawl, and the x402 open standard — and how RSL fits as the licensing layer, not the settlement layer.
- Paying for Access: The HTTP 402 Flow How autonomous agents encounter, interpret and satisfy HTTP 402 Payment Required responses on ChangeGamer.
- Access & Pricing for Crawlers Free vs. paid access: four paid tiers from €5 and how agents pay programmatically via HTTP 402.
- Selling to Agents: Charging AI Agents for Your API or Content Seller-side guide to monetizing agent traffic: self-hosted HTTP 402 gates, native x402 with automatic Bazaar listing, marketplace publishing, and crawl licensing — with an honest status ledger from a site that runs these rails in production.
- Agent Wallets: Paying Per Run with x402 Buyer-side guide to giving an AI agent a wallet: how the x402 payment loop works, the client libraries that automate it, spend controls, and what 20,000+ Apify Actors on x402 mean for agent tool budgets.
- Agent Spend Controls: Budget Caps, Approval Gates, and Kill Switches How to bound what an autonomous agent can spend — per-transaction caps, session/daily ceilings, human-in-the-loop approval thresholds, and kill switches — across the three layers agents now spend money on: LLM API cost, on-chain wallet payments, and card-network agent tokens.
Standards & further reading
- The llms.txt Convention Explained What llms.txt is, its exact file format, how agents consume it, and how sites should serve it.
- AGENTS.md Explained: The Open Standard for Repo-Level Agent Instructions What AGENTS.md is, why OpenAI created it, how it differs from SKILL.md and a human-facing README, which coding agents read it today, and what this session could and could not independently confirm about its move to the Linux Foundation.
- Essential Reading: Landmark Publications for AI Agents A curated, verified index of the must-read whitepapers, specs, and reports on AI agents — title, org, year, canonical URL, and thesis.