Meta-skill for improving and optimizing prompts using Anthropic's prompt engineering best practices. Provides the 4-step improvement workflow (example identification, initial draft, chain of thought refinement, example enhancement), keyword registries for documentation lookup, and decision trees for improvement strategies. Use when improving prompts, optimizing for accuracy, adding chain of thought reasoning, structuring with XML tags, enhancing examples, or iterating on prompt quality. Delegates to docs-management skill for official prompt engineering documentation.
Transforms basic prompts into high-performance templates using Anthropic's 4-step workflow: example identification, structured drafting, chain-of-thought refinement, and example enhancement. Use when optimizing prompts for accuracy, adding reasoning steps, structuring with XML tags, or iterating on quality—always triggers docs-management first to query official prompt engineering documentation.
/plugin marketplace add melodic-software/claude-code-plugins/plugin install claude-code-observability@melodic-softwareThis skill is limited to using the following tools:
references/examples/advanced-transformations.mdreferences/examples/basic-transformations.mdreferences/examples/domain-specific.mdreferences/metadata/keyword-registry.mdreferences/metadata/tag-reference.mdreferences/patterns/cot-patterns.mdreferences/patterns/example-enrichment-patterns.mdreferences/patterns/prefill-patterns.mdreferences/patterns/system-prompt-patterns.mdreferences/patterns/xml-tagging-patterns.mdreferences/troubleshooting/common-issues.mdreferences/troubleshooting/debugging-guide.mdreferences/troubleshooting/tradeoffs-guide.mdreferences/workflows/improvement-workflow.mdreferences/workflows/iterative-refinement.mdreferences/workflows/test-case-generation.mdSTOP - Before improving ANY prompt, you MUST invoke the docs-management skill.
This is NOT optional guidance - it is a required execution step.
Invoke the docs-management skill BEFORE making any improvements:
Search for relevant documentation using natural language:
Query at least ONE of these topics based on your improvement needs:
Before improving a prompt, verify:
If ANY checkbox is unchecked, STOP and invoke the docs-management skill first.
This meta-skill replicates Anthropic's Console/Workbench prompt improver functionality within Claude Code. It provides workflows, patterns, and keyword registries to transform basic prompts into high-performance structured templates.
What this skill provides:
What this skill does NOT provide:
docs-management skill)docs-management for current guidance)Use this skill when:
What do you want to do?
The prompt improver enhances prompts through a structured 4-step process:
<instructions> - Task definition and behavioral guidelines<context> - Background information and relevant details<examples> - Demonstration cases<formatting> - Desired output format specification<thinking>, <analysis>, <answer><thinking> steps within examples showing intermediate reasoningDetailed workflow: references/workflows/improvement-workflow.md
An improved prompt typically includes:
<instructions>
Your task definition and behavioral guidelines
</instructions>
<context>
Background information and relevant details
</context>
<examples>
<example>
<input>Sample input</input>
<thinking>Step-by-step reasoning</thinking>
<output>Expected output</output>
</example>
</examples>
<formatting>
Specify desired output format and structure
</formatting>
Use these keywords to query the docs-management skill for official documentation:
| Topic | Query Keywords |
|---|---|
| Prompt Improver | prompt improver, prompt improvement, optimize prompts |
| Chain of Thought | chain of thought, CoT, thinking, step-by-step, reasoning |
| XML Tags | XML tags, structure prompts, tagging, XML structure |
| Examples | multishot, few-shot, example formatting, multishot prompting |
| System Prompts | system prompt, role, persona, Claude role |
| Clarity | clear, direct, explicit, specific instructions, be clear and direct |
| Claude 4.x | Claude 4, Claude 4.5, Sonnet 4.5, Opus 4.5, best practices |
| Topic | Query Keywords |
|---|---|
| Prefilling | prefill, response prefill, assistant prefill, output format |
| Long Context | long context, long context tips, document placement |
| Extended Thinking | extended thinking, thinking budget, deep reasoning |
| Prompt Chaining | chain prompts, prompt chaining, multi-step prompts |
Find documentation about chain of thought prompting and thinking tags
Find documentation about XML tags for structuring prompts
Find documentation about multishot prompting and example formatting
Find documentation about Claude 4 and Claude 4.5 prompting best practices
Find documentation about prefilling Claude's response for output control
Based on Anthropic's testing, prompt improvement typically yields:
Important: Improved prompts produce longer, more thorough responses. Consider trade-offs for latency-sensitive or cost-sensitive applications.
Load these files based on your specific needs:
/improve-prompt command--feedback for iterative refinement--generate-examples for prompts lacking examplesDate: 2025-12-03 Model: claude-opus-4-5-20251101
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.