The canonical vendor guide: clarity, examples, structure and chain-of-thought, straight from the lab.
Prompt engineering overview
Anthropic Nov 2024
Open platform.claude.com →Good research prompts share four parts: a persona (act as an equity analyst), a precise task with scope and geography, the context you already have, and a required output format. For deep research runs, add guardrails: which sources to prefer, what to exclude, and an instruction to flag uncertainty instead of guessing. Iterate; the second prompt informed by the first answer is where the quality jumps.
Part of AI search on eChai Explore.
On the AI map, 29 October 2026 AI assistants AI browsers Research and search
18 resources.
The canonical vendor guide: clarity, examples, structure and chain-of-thought, straight from the lab.
Anthropic Nov 2024
Open platform.claude.com →Updated best practices that transfer well to every frontier model, not just Claude.
Anthropic Nov 2025
Open platform.claude.com →Nine hands-on chapters with exercises; the fastest structured way to level up.
Anthropic Aug 2024
Open github.com →The free reference 3M+ people learned from, covering zero-shot to ReAct with examples.
DAIR.AI Jan 2025
Open promptingguide.ai →The underlying repo with papers, notebooks and lessons for going deeper.
DAIR.AI Feb 2025
Open github.com →A self-paced course version if you prefer structure over docs.
DAIR.AI Mar 2025
Open academy.dair.ai →Google's persona-task-context-format formula with role-specific examples; effective prompts average 21 words.
Google Oct 2024
Open services.google.com →The interactive version of the handbook, organised by job and use case.
Google Workspace Jan 2025
Open workspace.google.com →A two-minute read of prompting basics to share with non-technical teammates.
Google Feb 2025
Open support.google.com →Names the 8 components of a research prompt including guardrails and anti-hallucination clauses.
FindSkill.ai Apr 2025
Open findskill.ai →A copy-paste template that turns a fuzzy founder question into an analyst brief.
Armen Berenson Mar 2025
Open armenberenson.com →The meta-prompting trick: have the model interview you before it researches.
Seth Wylie Feb 2025
Open linkedin.com →Worked examples of weak vs strong research prompts with the outputs compared.
Ali Azimi Darmian Mar 2025
Open medium.com →OpenAI's own training material on briefing the research agent properly.
OpenAI Academy May 2025
Open academy.openai.com →An educator's battle-tested prompt patterns for rigorous, citable research output.
José Bowen (Teaching Naked) Apr 2025
Open teachingnaked.com →Shows how a standing system prompt about you and your company upgrades every future research answer.
Tal Raviv, Lenny's Newsletter Feb 2025
Open lennysnewsletter.com →A gentle training plan that builds daily prompting muscle instead of one-off hacks.
Hilary Gridley, Lenny's Newsletter Apr 2025
Open lennysnewsletter.com →Prompting techniques specifically aimed at getting fewer made-up answers.
IBM Jun 2024
Open mediacenter.ibm.com →The same ground, over in Understand your customers, our Starting Up track.