Academy
Course 401Advanced

Agents in Production

Design agents that are as simple as they can be, manage their context deliberately, red-team them before attackers do, deploy them with a feedback loop, connect them over MCP, and know the rules that apply in 2026.

A glowing glass octahedron connected by red light lines to six smaller glass nodes

Lessons

  1. 1. Anatomy of an agent

    An agent is a model using tools in a loop. Most products need a workflow; reach for full autonomy only when the path can't be scripted.

    9 min

  2. 2. Context engineering

    In 2026, the craft has moved from writing one good prompt to curating everything an agent sees over hundreds of turns.

    9 min

  3. 3. Red-teaming and agent security

    Once an agent reads untrusted content and holds real permissions, every webpage and document is a possible attacker.

    10 min

  4. 4. Deploy, monitor and learn from feedback

    Shipping is the start of training, not the end. Serve behind a standard API, watch the golden signals, and feed real outcomes back in.

    8 min

  5. 5. Connecting agents with the API and MCP

    The Model Context Protocol is how agents discover and call tools across products. Its 2026 revision made it stateless and simpler to host.

    8 min

  6. 6. Governance and risk in 2026

    Regulation is now in force, not on the horizon. Know the frameworks and the dates, and build the evidence as you train.

    8 min

Sources

Every source cited across this course's lessons.

  1. [1]Yao, S. et al. (2022). ReAct: Synergizing Reasoning and Acting in Language Models. ICLR 2023 · arXiv:2210.03629.
  2. [2]Schick, T. et al. (2023). Toolformer: Language Models Can Teach Themselves to Use Tools. NeurIPS 2023 · arXiv:2302.04761.
  3. [3]Shinn, N. et al. (2023). Reflexion: Language Agents with Verbal Reinforcement Learning. NeurIPS 2023 · arXiv:2303.11366.
  4. [4]Anthropic (2024). Building effective agents. Anthropic Engineering.
  5. [5]Liu, N. F. et al. (2023). Lost in the Middle: How Language Models Use Long Contexts. TACL 2024 · arXiv:2307.03172.
  6. [6]Anthropic (2025). Effective context engineering for AI agents. Anthropic Engineering, September 2025.
  7. [7]OWASP GenAI Security Project (2025). OWASP Top 10 for LLM Applications. genai.owasp.org.
  8. [8]Greshake, K. et al. (2023). Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection. AISec 2023 · arXiv:2302.12173.
  9. [9]Zou, A. et al. (2023). Universal and Transferable Adversarial Attacks on Aligned Language Models. arXiv:2307.15043.
  10. [10]OWASP GenAI Security Project (2025). OWASP Top 10 for Agentic Applications for 2026. genai.owasp.org, December 2025.
  11. [11]Perez, E. et al. (2022). Red Teaming Language Models with Language Models. EMNLP 2022 · arXiv:2202.03286.
  12. [12]OpenAI (2025). Chat Completions API reference. platform.openai.com.
  13. [13]Beyer, B. et al. (eds.) (2016). Site Reliability Engineering, ch. 6: Monitoring Distributed Systems. O'Reilly / Google.
  14. [14]Gao, L., Schulman, J. & Hilton, J. (2022). Scaling Laws for Reward Model Overoptimization. arXiv:2210.10760.
  15. [15]Anthropic (2026). Demystifying evals for AI agents. Anthropic Engineering, January 2026.
  16. [16]Kojen (2026). Kojen developer docs: REST API, MCP server and endpoints. kojen.ai.
  17. [17]Model Context Protocol (2025). Model Context Protocol Specification, revision 2026-07-28. modelcontextprotocol.io.
  18. [18]National Institute of Standards and Technology (2023). AI Risk Management Framework (AI RMF 1.0). NIST AI 100-1.
  19. [19]Future of Life Institute (2026). EU AI Act implementation timeline (updated 31 August 2026). artificialintelligenceact.eu.
  20. [20]Gebru, T. et al. (2018). Datasheets for Datasets. Communications of the ACM · arXiv:1803.09010.
  21. [21]Mitchell, M. et al. (2019). Model Cards for Model Reporting. FAT* 2019 · arXiv:1810.03993.