Anthropic posts free prompting playbook

Anthropic has made a session called “The prompting playbook” available through its Code with Claude event site after presenting it in London on May 19, 2026. The session page says the talk covers “how to apply core prompting principles to agentic systems that plan, act, and adapt,” and lists Margot van Laar, a member of technical staff at Anthropic, as the speaker. (claude.com) The release adds to Anthropic’s growing set of public training materials for developers and business users. (claude.com) Anthropic’s learning hub says the company offers courses, API guides and deployment best practices through Anthropic Academy, alongside resources on prompt design and implementation. ### Where did this playbook actually appear? Anthropic published the session through its Code with Claude 2026 event page rather than as a standalone blog post. The page lists “The prompting playbook” as a breakout session in London, scheduled from 16:50 to 17:35 on May 19, 2026, with a “Watch recording” link on the event site. (claude.com) The same event schedule places the session alongside workshops on managed agents, Claude Platform usage and production agent workflows. That positioning suggests Anthropic presented the playbook as part of a broader developer education push around building and operating Claude-based systems. (anthropic.com) ### What does Anthropic say the session covers? Anthropic’s event page describes the session in one sentence: “How to apply core prompting principles to agentic systems that plan, act, and adapt.” The published description does not provide a longer outline on the page that was publicly accessible through search results reviewed Sunday. (claude.com) Anthropic has separately published training material focused on prompt quality and testing. A video on Anthropic’s YouTube channel titled “Evaluate prompts in the Anthropic Console” says users can generate, test and evaluate prompts, including automatic test cases and output comparisons. (claude.com) ### Why does this fit Anthropic’s broader developer push? Anthropic’s learning hub says it offers courses on API development, Claude Code, business deployment and Model Context Protocol, with certificates for some courses. The site describes the academy as a place for “resources,” “tutorials,” and “best practices” for working with Claude. (claude.com) Anthropic has also been publishing adjacent guidance on agent design and context management. Its research page on building effective agents discusses frameworks, tool use and implementation choices, while an engineering post on context engineering describes that work as a progression from prompt engineering toward managing the full set of information used during inference. (youtube.com) ### Is this Anthropic’s first public material on prompting? Anthropic has released earlier prompting material, including a “Prompting 101” session published on YouTube after a May 22, 2025 event in San Francisco. That video featured Hannah Moran and Christian Ryan from Anthropic’s applied AI team. (anthropic.com) Anthropic also published a longer YouTube discussion on prompt engineering about a year earlier featuring Amanda Askell, Alex Albert, David Hershey and Zack Witten. That video focused on how prompt engineering had evolved and offered practical guidance for enterprise, research and general-use prompting. (anthropic.com) ### What can readers watch next? Anthropic’s Code with Claude page shows “The prompting playbook” alongside other May 19 sessions in London, including talks on Claude Managed Agents, Claude Platform usage and enterprise-scale AI systems. Anthropic’s Academy page separately links to courses on Claude Code, API development and workplace deployment. (youtube.com) Anthropic’s public YouTube channel also hosts related developer videos, including “Evaluate prompts in the Anthropic Console” and earlier prompting sessions. Those materials remain available as of May 24, 2026. (youtube.com) (claude.com) (youtube.com)