Search the docs, videos, blog, apps, teams, skills and release notes.
Open source. macOS, Windows and Linux.
At a glance
What it is
Advanced prompting techniques for any AI model covering system prompts, chain-of-thought reasoning, few-shot examples, structured output formats, context management, multi-turn conversation strategies, and a reusable prompt template library.
A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.
The whole file, exactly as it downloads. Read it before you rely on it.
SKILL.md
421 lines · 1,944 words
Use this skill when:
User asks about ai prompt crafter techniques or best practices
User needs guidance on ai prompt crafter concepts
User wants to implement or improve their approach to ai prompt crafter
Advanced prompting techniques for any AI model covering system prompts, chain-of-thought reasoning, few-shot examples, structured output formats, context management, multi-turn conversation strategies, and a reusable prompt template library.
Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.
Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.
Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.
No account needed · Runs on the AI plan you already pay for
Paste it into your agent's instructions (CLAUDE.md, AGENTS.md or a custom GPT), or add it to a bot in Murage.
Ask Ember for the bot, then switch the skill on in its settings.
Step 1
Send your Chief of Staff this message. It proposes the bot and its role first, and creates it when you say go.
Send this to Ember
“Create a bot that uses the AI Prompt Crafter skill, and show me the plan first.”
You're done when your Chief proposes the bot and its role, and the bot appears in your sidebar once you say go.
Step 2
You're done when you see the bot’s skills and the library.
Step 3
You're done when you have read what Skill Guard found and the skill shows as on for that bot.
Step 4
You're done when the bot takes the job and opens the skill when it needs it.
Prefer to do it yourself? Choose + in the sidebar, then New Bot.Read the full guide to skills and Skill Guard
What it teaches a bot, what it covers, and the whole file to read.
What it covers
When to Use
Questions to Ask First
Prompting Fundamentals
System Prompts
Chain-of-Thought Prompting
Few-Shot Prompting
Structured Output
Context Management
Multi-Turn Conversation Strategies
Prompt Template Library
Advanced Techniques
Troubleshooting Common Issues
Process
Output Format
Edge Cases
Example
The request falls outside the scope of ai prompt crafter
User needs a different specialized skill for their specific situation
The topic requires professional consultation beyond general guidance
Before crafting any prompt strategy, establish context:
What AI model(s) are you primarily working with? (ChatGPT, Claude, Gemini, Llama, Mistral)
What is the primary use case? (Writing, coding, analysis, creative, research, automation)
What is your technical comfort level? (Beginner, intermediate, power user, developer)
Are you prompting through a chat interface, API, or integrated tool?
What quality bar do you need? (Quick drafts vs publication-ready vs mission-critical)
Are you working with sensitive or confidential information?
Do you need consistent, repeatable outputs or creative variety?
What has frustrated you about AI outputs so far? Where does it fail?
Every strong prompt contains these elements:
Context: Background information the AI needs
Role: Who the AI should act as
Action: What specifically you want it to do
Format: How the output should be structured
Tone: The voice and style of the response
BAD: "Write me a marketing email"
BETTER: "Write a marketing email for our SaaS product launch"
GOOD: "Write a marketing email announcing our project management
tool to small business owners. Tone: professional but
friendly. Include a subject line, 3 key benefits,
and a CTA to start a free trial. Keep under 200 words."
GREAT: [Includes all of the above plus a system prompt with
brand voice guidelines, 2 examples of previous emails
that performed well, and specific constraints]System prompts set the persistent context and behavior for an AI conversation. They act as a "configuration file" that shapes every response.
You are [role/identity].
Your expertise includes:
- [Skill/knowledge area 1]
- [Skill/knowledge area 2]
Your communication style:
- [Tone characteristic]
- [Format preference]
Rules you must follow:
- [Constraint 1]
- [Constraint 2]
When you don't know something:
- [How to handle uncertainty]
Context about the user/project:
- [Relevant background information]You are a senior technical writer with 15 years of experience
in software documentation. You write for developer audiences.
Style: Active voice. One idea per sentence. Define acronyms
on first use. Use code blocks for code references. Prefer
concrete examples over abstract explanations.
Format: H2/H3 headers for structure. Code examples for any
technical concept. "Note:" callouts for warnings. Paragraphs
under 4 sentences. If ambiguous, ask clarifying questions.Chain-of-thought (CoT) prompting asks the AI to show its reasoning step by step before arriving at a conclusion. This dramatically improves accuracy on complex tasks.
Mathematical or logical reasoning
Multi-step analysis or complex decision-making
Debugging and troubleshooting
Comparing multiple options
Explicit CoT:
Analyze whether this business should enter the European market.
Think through this step by step:
1. Assess current market conditions in Europe for this product
2. Evaluate the company's readiness (resources, team, product-market fit)
3. Identify the top 3 risks and mitigations
4. Estimate financial implications (costs, timeline to profitability)
5. Give your recommendation with a confidence level
Show your reasoning for each step.Zero-shot CoT (simple trigger):
[Your question or task]
Let's work through this step by step.Structured CoT:
Organize your thinking into:
- OBSERVATIONS: What do we know from the given information?
- ASSUMPTIONS: What are we assuming?
- ANALYSIS: What does the evidence suggest?
- CONCLUSION: What is the answer, and how confident are we?Few-shot prompting provides the AI with examples of the desired input-output pattern before asking it to perform the task. One of the most powerful techniques for consistent, formatted outputs.
I need you to [task description]. Here are examples:
Example 1:
Input: [example input 1]
Output: [example output 1]
Example 2:
Input: [example input 2]
Output: [example output 2]
Now do the same for:
Input: [your actual input]
Output:Use 2-5 examples (diminishing returns after 5)
Cover edge cases (don't just show the happy path)
Keep examples consistent in format and quality
Order matters: Put the most representative example last
Vary your examples: Show different scenarios to help the AI generalize
JSON Output:
Analyze this product review and return as JSON:
{
"sentiment": "positive" | "negative" | "mixed",
"confidence": 0.0-1.0,
"key_themes": ["theme1", "theme2"],
"summary": "one sentence summary"
}
Review: [paste review text]Table Output:
Compare these 5 tools across these dimensions.
Return as a markdown table:
Columns: Tool Name | Free Tier | Best For | Key Limitation | Rating (1-5)Length: "Keep your response under 200 words"
Format: "Use markdown headers, bullets, and code blocks"
Structure: "Use the following template: [template]"
Language: "Write at an 8th-grade reading level"
Exclusion: "Do not include disclaimers or caveats"
Inclusion: "Must include at least 3 specific examples"Every AI model has a limited context window. Managing context effectively is critical for complex tasks.
HIGHEST PRIORITY (always include):
- The specific task/question
- Critical constraints and requirements
- Most relevant examples
HIGH PRIORITY (include when space allows):
- Background context
- Previous relevant conversation
MEDIUM/LOW PRIORITY (summarize or omit):
- Supplementary information, edge cases
- General knowledge the AI already hasChunking: Split into logical sections, process each independently, summarize, then synthesize
Progressive summarization: Summarize long doc -> identify key arguments -> drill into specific sections
Reference and retrieve: Share document in sections, have AI extract key facts, then ask questions
Refinement Loop:
Turn 1: Initial generation (broad prompt)
Turn 2: "Expand section 3 with more detail"
Turn 3: "Make the tone more conversational"
Turn 4: "Add a conclusion that ties back to the opening"
Turn 5: "Fix any inconsistencies and polish"Interview Method:
Turn 1: "I need help with [topic]. Before you start, ask me
5 questions that will help you give better advice."
Turn 2: [Answer the AI's questions]
Turn 3: "Based on my answers, provide your recommendations."Periodically recap: "To summarize where we are: [summary]"
Reset context: "Set aside the tangent about X. Refocus on Y."
Be explicit: "In your last response, you mentioned [point]. Expand on that."
Number iterations: "This is revision 3. Here's what still needs work..."
Role: Experienced content creator and strategist.
Task: Create a [content type] about [topic] for [audience].
Tone: [professional / casual / authoritative / friendly]
Length: [word count] Platform: [where published]
Goal: [what this should achieve] Keywords: [list]
Examples of similar content I've liked: [optional]Language: [language] Framework: [if applicable]
Task: [what the code should do]
Requirements: [functional requirements]
Constraints: [performance, security, style]
Include: Code with comments, brief explanation, example usage, caveatsAnalyze [subject] with focus on: [objective]
Framework: [specific framework or lenses]
Deliverable: Executive summary, detailed analysis, key findings
(ranked), recommendations with confidence levels, risks/caveatsUse the output of one prompt as input for the next:
Chain 1: "Generate 10 blog post title ideas about [topic]"
Chain 2: "For title #3, create a detailed outline"
Chain 3: "Write section 2 of this outline in full"
Chain 4: "Review and suggest improvements""You are three experts debating this topic:
- Expert A: A skeptical data scientist who wants evidence
- Expert B: A creative marketer who thinks about narrative
- Expert C: A pragmatic CEO who cares about ROI
Have them discuss [topic] and reach a consensus recommendation.""I need to create a prompt for [task]. Generate a prompt that
includes all necessary context, uses the most effective technique,
specifies output format, and includes relevant constraints.
Also explain why you chose this approach."Tell the AI what NOT to do:
Do NOT: Use cliches, start with "In today's digital age...",
include generic disclaimers, use passive voice, pad with
unnecessary words, or list obvious information.| Problem | Likely Cause | Fix |
|---|---|---|
| Output too generic | Prompt too vague | Add specific context, examples, constraints |
| Output too long | No length constraint | Specify word count or section limits |
| Inconsistent format | No format specification | Provide a template or examples |
| Hallucinated facts | Requires specific knowledge | Ask AI to flag uncertainty |
| Ignoring instructions | Too many instructions | Prioritize and number instructions |
| Wrong tone | No tone guidance | Include explicit tone instructions with example |
| Shallow analysis | No depth instruction | Ask for step-by-step reasoning |
Gather information. Ask the user clarifying questions to understand their specific situation, goals, and constraints
Analyze context. Review the information provided and identify key factors relevant to ai prompt crafter
Develop recommendations. Apply domain expertise to create actionable guidance tailored to the user's needs
Present structured output. Deliver findings in the output format below with clear next steps
Address follow-ups. Answer additional questions and refine recommendations based on feedback
## Ai Prompt Crafter Analysis
### Assessment
[Key findings and observations]
### Recommendations
1. [Primary recommendation]
2. [Secondary recommendation]
3. [Additional suggestions]
### Action Items
- [ ] [First action step]
- [ ] [Second action step]
- [ ] [Follow-up task]Incomplete information: Ask clarifying questions before proceeding with recommendations
Conflicting requirements: Prioritize the most critical constraint and note trade-offs
Out of scope requests: Redirect to appropriate specialized skill or professional resource
Beginner vs advanced: Adjust depth and terminology based on user's experience level
Input: "Help me with ai prompt crafter for my current situation"
Output:
Based on your situation, here is a structured approach to ai prompt crafter:
Assessment: Evaluate your current state and identify key areas for improvement
Strategy: Develop a targeted plan based on best practices
Implementation: Execute the plan with specific, measurable steps
Review: Monitor progress and adjust as needed
SKILL.md
421 lines · 1,944 words
The text of SKILL.md is shown as it downloads, under the Apache-2.0 license, by Ferrox Labs. The license is the one in the file's own frontmatter.
Your new bot
AI Prompt Crafter bot
Waiting in your sidebar once you say go.
No account needed · Runs on the AI plan you already pay for