{"title":"Prompt \u0026 Agent Packs","description":"\u003cp\u003eReady-to-use system prompts, agent instruction sets, and rules files you can paste in and use today.\u003c\/p\u003e","products":[{"product_id":"the-agent-instruction-pack-40-system-prompts-for-real-work","title":"The Agent Instruction Pack: 40 System Prompts for Real Work","description":"\u003cp\u003eMost prompt packs are lists of one-liners that fall apart the moment you use them on real work. This one is different. The Agent Instruction Pack gives you 40 complete system prompts for the jobs developers, creators, operators, and indie hackers actually do every week: reviewing pull requests, debugging, grading research sources, editing without losing your voice, turning meeting notes into owned action items, triaging an inbox, and answering support tickets without inventing policy. Each one is built the same way: a clear role, fill-in-the-blank [VARIABLES], step-by-step instructions, guardrails so your agents stay on task, make fewer things up, and don't make promises, a defined plan for when it's uncertain, and an exact output format.\u003c\/p\u003e\n\u003cp\u003eEvery prompt comes with when to use it, a guardrails-at-a-glance summary, a short example input with the output shape to expect, and a practical pro tip. They're model-agnostic, so you can paste them into ChatGPT, Claude, Grok, Gemini, a custom GPT, a Project, or your own agent's system message. You get a 93-page PDF designed for reading and a ZIP of Markdown files (one per prompt, in category folders) that you can drop into Notion, Obsidian, a repo, or your prompt library.\u003c\/p\u003e\n\u003ch3\u003eWhat's included\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003e40 complete system prompts\u003c\/strong\u003e in 5 categories of 8: Coding, Research, Writing, Operations, and Support\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eBuilt-in guardrails on every prompt:\u003c\/strong\u003e what it must never do, what to do when it's unsure, and the exact output format\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e[BRACKETED] variables\u003c\/strong\u003e so you can adapt any prompt in under a minute, plus {flag} markers that show you what to check before you ship\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eExample input + expected output shape\u003c\/strong\u003e for every prompt, so you know what \"good\" looks like\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e\"How to adapt these to your role\" guide\u003c\/strong\u003e (3 pages): making prompts permanent, adding rules from your own failures, chaining prompts into workflows, and using them safely in agents and automations\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eOne-page prompt-structure cheat sheet:\u003c\/strong\u003e the six-block structure, guardrail phrases worth stealing, and a prompt-debugging table\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e93-page PDF\u003c\/strong\u003e with a clickable table of contents\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eZIP of 44 Markdown files\u003c\/strong\u003e, organized in category folders and ready for Notion, Obsidian, GitHub, or any prompt manager\u003c\/li\u003e\n\u003cli\u003eWorks with ChatGPT, Claude, Grok, Gemini, custom GPTs, and agent frameworks\u003c\/li\u003e\n\u003cli\u003eInstant download. Personal and commercial use in your own work.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cstrong\u003eThe 40 prompts:\u003c\/strong\u003e\u003cbr\u003e\nCoding: Code Reviewer · Debugging Partner · Test Writer · Refactor Planner · PR Description Writer · Migration Assistant · API Documentation Writer · Security Reviewer\u003cbr\u003e\nResearch: Source-Grading Researcher · Competitor Teardown · Market Sizing Analyst · Literature Summarizer · Fact-Checker · Customer Interview Synthesizer · Tool \u0026amp; Vendor Evaluator · Decision Brief Writer\u003cbr\u003e\nWriting: Voice-Preserving Editor · Newsletter Drafter · Landing Page Copywriter · Technical Explainer · Thread Repurposer · Cold Outreach Writer · Headline \u0026amp; Hook Generator · Release Notes Writer\u003cbr\u003e\nOperations: Meeting Notes to Action Items · SOP Writer · Inbox Triage Assistant · Weekly Status Report · Project Risk Scanner · Blameless Postmortem Writer · Hiring Scorecard Builder · Automation Opportunity Mapper\u003cbr\u003e\nSupport: Support Reply Drafter · Escalation Classifier · FAQ Builder · Refund Policy Enforcer · Bug Report Translator · Cancellation Save Responder · Support QA Grader · Voice-of-Customer Reporter\u003c\/p\u003e","brand":"Prompt \u0026 Process","offers":[{"title":"Default Title","offer_id":48282609713323,"sku":null,"price":29.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0804\/1535\/9147\/files\/cover-2048.png?v=1792157844"},{"product_id":"the-coding-agent-rules-pack-drop-in-rules-for-cursor-claude-code-codex-agents-md-and-github-copilot","title":"The Coding Agent Rules Pack: Drop-In Rules for Cursor, Claude Code, Codex (AGENTS.md) and GitHub Copilot","description":"\u003cp\u003eYour coding agent is quick. It is also happy to announce \"done\" without running the tests, rewrite a whole file to change one line, and install a package that doesn't exist. The Coding Agent Rules Pack gives it house rules: ready-to-drop-in instruction files that tell it to run your real commands and show the results before claiming done, keep diffs small, check that every package exists before installing it, keep secrets out of code and commits, follow your code style and git habits, and ask instead of guessing when something is hard to undo.\u003c\/p\u003e\n\u003cp\u003eThe files are organized by tool, in each tool's own format: \u003ccode\u003e.cursor\/rules\/*.mdc\u003c\/code\u003e with proper frontmatter for Cursor, \u003ccode\u003eCLAUDE.md\u003c\/code\u003e plus \u003ccode\u003e.claude\/rules\/\u003c\/code\u003e for Claude Code, \u003ccode\u003eAGENTS.md\u003c\/code\u003e for OpenAI Codex and other AGENTS.md-aware agents, and \u003ccode\u003e.github\/copilot-instructions.md\u003c\/code\u003e for GitHub Copilot. Pick your stack, copy one folder, fill in a short project-context template, and you're set up. The 18-page PDF guide explains how each tool loads rules, how to install, customize, and combine them, the mistakes that make rules fail silently, and ends with a one-page cheat sheet. Every file path and format was checked against each tool's official docs in October 2026, and the guide lists the sources so you can re-check them when the tools change.\u003c\/p\u003e\n\u003ch3\u003eWhat's included\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eUniversal base rules:\u003c\/strong\u003e read before editing, ask when unsure, small diffs, dependency and package-hallucination checks, no secrets, code style, verify before \"done\", git hygiene, and a standard end-of-task summary\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e8 stack variants\u003c\/strong\u003e with real commands and stack-specific traps: TypeScript + Next.js, React + Vite, Node.js + Express API, Python + FastAPI, Python data and notebooks, Go service, Rust, React Native (Expo or bare)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e6 task-mode add-ons:\u003c\/strong\u003e bugfix, refactor, tests-first, security review, docs, and always-on \"don't touch\" boundaries\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eFill-in Project Context template:\u003c\/strong\u003e commands, layout, conventions, environment, owners, gotchas\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCursor set:\u003c\/strong\u003e \u003ccode\u003e.cursor\/rules\/*.mdc\u003c\/code\u003e files with \u003ccode\u003ealwaysApply\u003c\/code\u003e \/ \u003ccode\u003edescription\u003c\/code\u003e frontmatter, plus 8 stack rules\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eClaude Code set:\u003c\/strong\u003e \u003ccode\u003eCLAUDE.md\u003c\/code\u003e, \u003ccode\u003e.claude\/rules\/\u003c\/code\u003e files, and mode files to @-mention\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eAGENTS.md set:\u003c\/strong\u003e ready-assembled \u003ccode\u003eAGENTS.md\u003c\/code\u003e for each stack plus a generic one\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eGitHub Copilot set:\u003c\/strong\u003e \u003ccode\u003e.github\/copilot-instructions.md\u003c\/code\u003e for each stack, a path-specific \u003ccode\u003etests.instructions.md\u003c\/code\u003e, and the task modes as \u003ccode\u003e.prompt.md\u003c\/code\u003e files\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eTool-neutral source modules\u003c\/strong\u003e so you can maintain one set of rules for several tools\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e18-page PDF guide\u003c\/strong\u003e with install commands, customization, combining files, common mistakes, troubleshooting, a one-page cheat sheet, and the official doc sources checked\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e83 files in one ZIP\u003c\/strong\u003e, plain Markdown, edit anything\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cstrong\u003eWritten for\u003c\/strong\u003e Cursor, Claude Code, OpenAI Codex, and GitHub Copilot as documented in October 2026. AI coding tools change their file conventions over time; the guide shows how to check that each file is loaded.\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eLicense:\u003c\/strong\u003e personal and commercial use in your own projects (including client and employer projects you work on). No resale or redistribution of the files.\u003c\/p\u003e\n\u003cp\u003eInstant download. Not affiliated with Cursor, Anthropic, OpenAI, or GitHub.\u003c\/p\u003e","brand":"Prompt \u0026 Process","offers":[{"title":"Default Title","offer_id":48282609746091,"sku":null,"price":24.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0804\/1535\/9147\/files\/cover_d002363e-4311-4d3f-9ae0-bb3115e19089.png?v=1792157887"}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0804\/1535\/9147\/collections\/cover-2048.png?v=1792158594","url":"https:\/\/promptandprocess.shop\/collections\/prompt-agent-packs.oembed","provider":"Prompt \u0026 Process","version":"1.0","type":"link"}