Sources

Source policy

This site prefers official framework documentation, standards bodies, security projects, and primary research. Framework comparisons are dated because product surfaces change. Vendor claims are used to explain capabilities, not to rank vendors. Practical recommendations come from matching those capabilities to engineering constraints.

The T4 treatment also publishes llms.txt and llms-full.txt so AI systems can discover page intent, canonical URLs, and source context without scraping the full site first.

Bibliography

1. Agents SDK
   OpenAI Developer Docs. Accessed 2026-07-06.
   Defines agents as applications that plan, call tools, collaborate, and keep state.

2. OpenAI Agents SDK
   OpenAI. Accessed 2026-07-06.
   Documents the Python Agents SDK primitives: agents, handoffs, guardrails, tracing, sessions, tools, and sandbox agents.

3. Guardrails
   OpenAI Agents SDK. Accessed 2026-07-06.
   Explains input, output, and tool guardrail boundaries in OpenAI Agents SDK workflows.

4. Tracing
   OpenAI Agents SDK. Accessed 2026-07-06.
   Describes built-in tracing for LLM generations, tool calls, handoffs, guardrails, and custom events.

5. Function Calling
   OpenAI Developer Docs. Accessed 2026-07-06.
   Official guide to exposing application functions and external systems to models through tools.

6. Building Effective Agents
   Anthropic Engineering. Accessed 2026-07-06.
   A practical distinction between augmented LLMs, workflows, and autonomous agents.

7. Tool Use with Claude
   Anthropic Platform Docs. Accessed 2026-07-06.
   Explains how Claude selects tool calls and how applications execute client-side tools.

8. Computer Use Tool
   Anthropic Platform Docs. Accessed 2026-07-06.
   Documents the browser/desktop agent loop and the security precautions for computer use.

9. Agent SDK Overview
   Claude Code Docs. Accessed 2026-07-06.
   Documents Claude Agent SDK built-in tools, hooks, subagents, MCP, permissions, sessions, and production use cases.

10. LangGraph Overview
    LangChain Docs. Accessed 2026-07-06.
    Positions LangGraph as an orchestration runtime for durable execution, streaming, HITL, and persistence.

11. Persistence
    LangChain Docs. Accessed 2026-07-06.
    Explains short-term memory through checkpointers and long-term memory through stores.

12. CrewAI Documentation
    CrewAI. Accessed 2026-07-06.
    Introduces CrewAI agents, crews, flows, guardrails, memory, knowledge, and observability.

13. Agents
    CrewAI. Accessed 2026-07-06.
    Defines CrewAI agents as autonomous units with roles, goals, tools, memory, and collaboration.

14. Flows
    CrewAI. Accessed 2026-07-06.
    Documents structured, event-driven workflows with state management, conditional logic, and loops.

15. ReAct: Synergizing Reasoning and Acting in Language Models
    arXiv. Accessed 2026-07-06.
    Research paper on interleaving reasoning traces and task-specific actions.

16. Toolformer: Language Models Can Teach Themselves to Use Tools
    arXiv. Accessed 2026-07-06.
    Research paper on models deciding which external API to call, when, and with which arguments.

17. Model Context Protocol Specification
    Model Context Protocol. Accessed 2026-07-06.
    Official specification for resources, prompts, and tools exposed by MCP servers.

18. AI Risk Management Framework
    NIST. Accessed 2026-07-06.
    NIST overview of AI RMF 1.0 and the 2024 Generative AI Profile.

19. AI RMF Core
    NIST AI Resource Center. Accessed 2026-07-06.
    Summarizes Govern, Map, Measure, and Manage as the AI RMF Core functions.

20. 2025 Top 10 Risk and Mitigations for LLMs and Gen AI Apps
    OWASP Gen AI Security Project. Accessed 2026-07-06.
    Current OWASP LLM risk categories including prompt injection, sensitive data, excessive agency, and output handling.

21. OWASP Top 10 for LLM Applications 2025
    OWASP Gen AI Security Project. Accessed 2026-07-06.
    Background resource on the Top 10 for Large Language Model Applications.