ChatGPT·16 min
Prompting ChatGPT Agent Mode (Tasks and Guardrails)
ChatGPT agent prompts need a stop condition and a scope, not just a goal. How agent mode's guardrails work, and what OpenAI's own docs say about ChatGPT Work replacing it.
Read moreBlog
Better prompts. Better AI results.
Tutorials, comparisons, and frameworks to write better prompts for ChatGPT, Claude, Gemini, Midjourney, Veo3 and Kling.
Page 9 of 29
Engineering·13 min
Pair Programming with AI (A Working Method)
A four-phase method for pair programming with AI — plan, implement, verify, review — plus what GitHub's and METR's own controlled studies actually measured about the speed gain.
Read moreEngineering·12 min
Step-Back Prompting (Ask the Bigger Question First)
Step back prompting has the model answer a bigger, more general question first, then reason from that to yours. The ICLR 2024 paper, the two-step method, and copy-paste templates.
Read moreEngineering·10 min
Analogical Prompting (Let the Model Generate Its Own Examples)
Analogical prompting has a model generate its own worked examples before answering, instead of you writing few-shot examples. The published technique, its own benchmark numbers, and a template.
Read moreEngineering·13 min
Does Instruction Order Change the Output? (Yes)
Prompt instruction order changes what a model does with conflicting or repeated instructions. What OpenAI's and Anthropic's own docs say, plus two peer-reviewed papers on prompt order sensitivity.
Read moreEngineering·12 min
Self-Ask Prompting (Let the Model Interrogate Itself)
Self ask prompting has a model break your question into its own follow-up questions and answer each one before answering yours. The paper that named it, its own benchmark numbers, and a template.
Read moreChatGPT·12 min
Prompting in Google AI Studio (For Developers)
Google AI Studio prompts: where system instructions, safety thresholds, and structured output actually live in the Run settings panel, and what changes when you turn one into Gemini API code.
Read moreChatGPT·13 min
The Same Prompt Across Six Models (Full Comparison)
Same prompt, different AI: a verified six-model comparison method, with a copy-paste test prompt, scoring rubric, and the documented reasons ChatGPT, Claude, Gemini, Grok, DeepSeek and Kimi diverge.
Read moreEngineering·13 min
Prompting for Performance Optimisation
Performance optimization prompts work when they force measurement first: a profile, a hypothesis, and a before/after benchmark, not a confident guess about what's slow.
Read moreEngineering·16 min
Prompting for SQL Optimisation
SQL optimization prompts that actually work: what to paste besides the query, how EXPLAIN differs across Postgres, MySQL and SQLite, and a copy-paste template.
Read moreChatGPT·13 min
Prompting Gemini Inside Google Workspace
Real gemini workspace prompts for Gmail, Docs, Sheets, and Slides, sourced to Google's own plan matrix, its official prompt framework, and the rules that repeat across every app. Dated and quoted.
Read moreChatGPT·13 min
Prompting for Legal and Medical Topics: The Hard Limits
Where AI legal medical prompts hit a hard wall: what OpenAI, Anthropic and Google policies actually say, what HIPAA and UK GDPR require, and the workflow to use instead.
Read moreEngineering·11 min
Grounding Techniques: Keeping AI Tied to Your Facts
Grounding AI prompts means restricting a model to text you provide, not its training data. Anthropic's own verified techniques for this, quoted directly, and where they still fail.
Read moreVideo·13 min
From Storyboard to Finished Sequence (Full Pipeline)
A full storyboard to video ai pipeline: structured shot prompts, locked characters and style, per-model first/last-frame handoff, extension, transitions, audio, and one export.
Read moreEngineering·12 min
Prompt Caching Parameters Across Providers
Prompt caching API parameters compared across Anthropic, OpenAI, and Gemini — cache_control, cached tokens, and context caching TTLs, minimums, and pricing, verified September 2026.
Read moreChatGPT·14 min
Provider Settings Compared (OpenAI vs Anthropic vs Google)
An api settings comparison of OpenAI, Anthropic and Google covering temperature, top_p, top_k, penalties, seed and stop sequences, checked against each vendor's live docs, Sept 3, 2026.
Read moreLibrary·12 min
Best Mobile Apps for Managing AI Prompts (iOS and Android)
An honest AI prompt app mobile shortlist for iOS and Android: which apps are real, which one is name-squatting a bigger brand, and why Prompt Architects still has no native app there.
Read moreIndustries·17 min
20 AI Prompts for Webinars and Live Events
20 copy-paste webinar ai prompts covering the full arc: promotion, registration, run-of-show, the live hour, and follow-up. Real prompts, not fill-in-the-blank stubs.
Read moreIndustries·17 min
20 AI Prompts for Risk Assessment and Pre-Mortems
20 risk assessment ai prompts covering identification, likelihood/impact scoring, Gary Klein's pre-mortem method, mitigation planning, and the review cadence that keeps a register alive.
Read moreIndustries·16 min
20 AI Prompts for Accessible, Plain-Language Copy
20 plain language AI prompts for auditing, rewriting, and verifying accessible copy — plus the WCAG checks a prompt alone can't pass.
Read moreChatGPT·13 min
Open-Weight vs Closed Models: What Changes in Your Prompt
What changes with open source model prompting versus a closed API: licensing, chat templates, system-prompt authority, and sampling controls — verified against vendor sources.
Read moreLibrary·14 min
Best AI Tools for Newsletter Writers (Prompts Edition)
A newsletter ai prompt tools comparison — ChatGPT, Claude, Jasper, beehiiv AI, and Prompt Architects — judged on whether your voice survives issue 50, not issue 1.
Read moreEngineering·15 min
Free Newsletter Prompt Generator (Subject to Signoff)
A free AI newsletter generator prompt that drafts a whole issue, subject line to signoff, matched to your own voice, plus the consistency check that keeps issue 40 sounding like issue 4.
Read moreIndustries·18 min
30 AI Prompts for X/Twitter Threads
30 copy-paste twitter thread prompts for the hook, the escalation, the payoff, each individual post, the visual, the CTA, and turning a blog post into a thread.
Read morePage 9 of 29