Custom Claude AI Rules Setup
$10-30 CAD
Posted 1 minute ago
•
Ends in 6 days
$10-30 CAD
Paid on delivery
I’m rolling out Anthropic’s Claude across several workflows and need a clear, well-structured set of skills and governing rules so the model answers reliably, stays on-brand, and avoids safety pitfalls. I’ll explain my use cases in a short kickoff call; from there you’ll translate those needs into system-level instructions, developer messages, and reusable prompt templates that slot straight into Claude’s API or chat interface. What I expect back: • A documented rulebook (system + developer messages) with comments explaining each guideline • Modular skill definitions or prompt snippets covering the scenarios we discuss • Example conversations showing the expected behaviour and tone, ready for quick A/B testing • A concise hand-off guide so I can adjust or extend the rules without breaking consistency I’ll consider the job complete once Claude consistently follows the provided rules, answers accurately, and handles edge cases we outline together. Familiarity with Anthropic’s safe-completion parameters, prompt engineering best practices, and iterative testing is essential.
Project ID: 40830516
About the project
16 proposals
Open for bidding
Remote project
Active 7 secs ago
Place your bid
Bid amount Email addressBenefits of bidding on Freelancer
Set your budget and timeframe
It's free to sign up and bid on jobs
16 freelancers are bidding on average $20 CAD for this job


Hello, After reviewing your project, I see you need a Claude rulebook and skill set that keeps outputs reliable, on-brand, and safe across your workflows. I’ve built similar governance layers for AI assistants, so I can translate your use cases into precise system prompts, developer messages, and reusable templates that plug directly into Claude’s API or chat. A practical approach here is to start with your kickoff call, define the guardrails (tone, forbidden topics, fallback behaviors), and then structure them as modular instructions that are easy to test A/B and extend later without breaking consistency. A couple of quick questions: • Are there specific compliance or safety standards (e.g., GDPR, internal policy) the rules must enforce? • Should the templates include few-shot examples for each scenario, or will system instructions be enough? I can deliver the documented rulebook, modular snippets, example conversations, and a hand-off guide within 3–4 days for CAD 25, given the scope described. I would be glad to discuss further details and am ready to start immediately. Looking forward to hearing from you. Best regards, Carlos.


Hi, I've spent the last decade building systems that need to behave consistently and predictably—APIs, automation workflows, rule engines—so I know how to translate business logic into clear, testable instructions. I'll start with a kickoff call to map your specific use cases, then build a documented rulebook with system prompts, developer messages, and reusable templates that integrate directly into Claude's API. Each rule will include comments explaining the reasoning, and I'll create example conversations for each scenario so you can verify behavior before rolling out. I'll test against your edge cases using Claude's safety parameters to make sure responses stay on-brand and handle exceptions without breaking. The handoff guide will be straightforward enough that you can adjust rules or add new skills without needing me to translate—you'll own the system from day one. I'm available now and will stay on for any refinements once you're live.


Hi there, I see that you're looking to roll out Claude effectively across various workflows while ensuring it adheres to specific guidelines and maintains brand consistency. I can help create a comprehensive rulebook that includes modular skill definitions and prompt templates tailored for the Anthropic API. My previous experience in prompt engineering ensures I can structure the requirements you have mentioned clearly and effectively. The project will involve documenting system-level instructions with detailed comments, producing example conversations for A/B testing, and providing a hand-off guide for future adjustments. My familiarity with best practices in prompt engineering will be leveraged to help Claude meet your outlined safety and accuracy benchmarks. I look forward to discussing the details and aligning on your specific needs. Regards, Ali


I can build a reliable automated solution and workflow for Custom Claude AI Rules Setup. I specialize in Python and n8n automation, building automated data processing pipelines that handle API requests, error retries, and data transformations into structured formats with zero downtime. Here is how I will handle this project: 1. Map out the exact data inputs, triggers, and expected outputs. 2. Develop and test the Python/n8n automation with error handling and validation. 3. Deliver fully documented code and setup instructions for seamless execution. I can start today and deliver within the agreed timeframe. Best regards, Bayush


Hi, getting Claude to answer reliably across several workflows comes down to clear, layered instructions: a solid system prompt for each workflow, reusable skills for repeated tasks, and explicit rules for tone, format and what to refuse or escalate. Then testing them against real examples before rollout. What I'll deliver after our kickoff call: - A system prompt for each workflow, defining role, audience, brand voice, output format and boundaries - Reusable skills and instruction files (e.g. Claude Projects instructions or a CLAUDE.md-style rule set) for recurring tasks such as summaries, replies and analysis, with examples of good output - Guardrails: what to never do, how to handle uncertainty ("say you don't know"), sensitive data rules, and when to hand over to a human - A test set of real prompts and edge cases for each workflow, run against the instructions, with fixes until the answers are consistent - A short guide for your team on using and updating the rules I have experience in this type of work: I work with AI tools and chatbot development daily, including prompt design and AI integrations through APIs. Best regards, Aryan


HI! IMPORTANT: CLARITY, CONSISTENCY, SAFETY Clear system instructions, consistent application of rules, and safety measures are crucial for successful implementation. Extensive experience with Anthropic’s Claude and prompt engineering has equipped me to mitigate risks and ensure reliable outputs. I will develop a structured rulebook, modular skill definitions, and example conversations based on your use cases, ensuring clarity and safety in responses. You will receive optimal results by leveraging the desired programming language and features tailored to your needs. I MISS YOUR RESPONSE! GOOD LUCK


Hello, I hope this message finds you well. I’m excited about the opportunity to collaborate on your project to refine Anthropic’s Claude for your specific workflows. Your goal to establish a robust framework for reliable, on-brand, and safe AI interactions is a challenge I’m well-equipped to tackle. With extensive experience in AI chatbot development and prompt engineering, I have a strong track record of creating custom rules and prompt templates that align with user needs while ensuring compliance and safety. My familiarity with Anthropic’s safe-completion parameters and iterative testing methods will be instrumental in delivering a solution that meets your standards. Here’s how I propose to achieve your objectives: First, I’ll conduct a kickoff call to thoroughly understand your use cases. Based on this, I will craft a comprehensive rulebook with detailed comments for clarity. This will include modular skill definitions and prompt snippets tailored to your scenarios. I’ll also provide example conversations, facilitating immediate A/B testing to ensure Claude’s responses align perfectly with your expectations. Finally, I’ll prepare a straightforward hand-off guide, empowering you to make future adjustments effortlessly while maintaining rule consistency. My goal is to ensure Claude’s performance is not only accurate but also adaptable to new situations and requirements. I’m looking forward to the possibility of working together to make Claude a seamless part of your operations. Best regards, Dominique


Hi, setting up custom rules and modular skills for Anthropic’s Claude to ensure reliable, on-brand, and safe responses fits my experience well. I’ve recently worked with Therabot fine-tuning Llama models with FastAPI backends and contributed to frontier labs evaluating AI-generated code, focusing on prompt engineering, fallback mechanisms, and safety. Here’s how I’d handle it: 1) define clear system-level instructions with safety guardrails using Anthropic’s safe-completion parameters to prevent undesired outputs; 2) craft reusable, modular prompt snippets and developer messages tailored to each workflow scenario for consistency; 3) deliver example conversations and a detailed guide for you to safely adjust rules while maintaining consistent behaviour. I estimate 3 to 4 weeks with milestones for kickoff and analysis, rulebook and prompts, and validation/testing; I can start now. Julia


Hello, I’m Kristián Lauko, with 10+ years of experience in AI Chatbot Development, AI Development, and prompt engineering for reliable assistant behavior. I understand your goal: to build a structured Claude setup with clear system and developer rules, reusable prompt templates, and consistent on-brand responses. I’ll turn your workflow needs into a documented rulebook, modular skill snippets, example conversations, and a practical hand-off guide that makes future edits straightforward while keeping behavior stable across edge cases. I have extensive experience designing instruction frameworks, safe-completion logic, and testing prompt behavior across complex use cases, so I can help Claude stay accurate, controlled, and aligned with your standards. I’d ensure effective communication throughout the project and I’m excited to hear more about your use cases. Thanks, Kristián Lauko


Hi, I'm a senior C#/.NET developer with a background in API integrations, backend architecture and reliable application workflows. I'd structure your Claude instructions into reusable rules, task-specific prompts and test cases, with clear boundaries for tone, accuracy and safety. I'd also provide A/B examples and a short guide for maintaining the rules. Are you using Claude through the API, Claude Projects, or both? Happy to discuss.


I can prepare a clear Claude rulebook with commented system and developer instructions, modular prompt snippets for your workflows, and example conversations for A/B testing. The hand-off guide will show how to extend the rules consistently and preserve your brand voice and safety expectations. After the kickoff, the deliverables can be organized around your use cases and the edge cases you identify, with an iterative review of the examples to refine expected behavior.


I’ve developed chatbot solutions focused on reliable, consistent responses, which fits the rulebook and reusable prompt work you need. For this project, the key is to separate stable system-level rules from workflow-specific developer instructions so updates stay scoped and safety guidance remains clear. I’ll turn your kickoff use cases into a commented rulebook, modular skill snippets, and example conversations that demonstrate the intended tone and edge-case behavior. I’ll also provide a concise guide for extending the materials, with example variants organized for A/B comparison in Claude. We can refine the examples against your acceptance criteria so the handoff reflects the behavior you want. Thanks, Kittipong


Hello, ensuring Claude’s responses align with your brand while navigating safety pitfalls is crucial. The key challenge will be translating your use cases into structured guidelines that maintain consistency across different contexts. I recently worked on a project where I developed a complex rulebook for an AI chatbot used in customer support. I created modular prompt templates and system messages that adapted to various scenarios while ensuring accurate and on-brand responses. This involved a lot of iterative testing and refining the prompts based on real user interactions. For your project, I would start by gathering all your use cases in our kickoff call, then draft the system-level instructions and developer messages. I’ll structure the documentation clearly, with comments explaining each guideline, and create example conversations for A/B testing. The goal is to provide a hand-off guide that allows you to easily adjust the rules without sacrificing consistency. Do you have any specific scenarios or edge cases in mind that we should prioritize? Looking forward to your thoughts. Thanks, Marcio.
内容来自 Freelancer News,版权归原作者所有。
阅读原文 ↗