Prepare book book_20260922_163508_a82865

books/book_20260922_163508_a82865/manifest.json

@@ -1,11 +1,58 @@
 {
+  "schema_version": "aiocr_book_manifest_v2",
   "book_id": "book_20260922_163508_a82865",
-  "book_name": "كتاب بدون عنوان",
+  "title": "كتاب بدون عنوان",
   "created_at": "2026-09-22T16:35:08.901520+00:00",
-  "source_files": [
-    ""
+  "text_repo_key": "aiocr_assistant",
+  "source_manifest": [
+    {
+      "source_index": 1,
+      "source_name": "The-Complete-Guide-to-Building-Skill-for-Claude.txt",
+      "source_type": "device_txt",
+      "char_start": 0,
+      "char_end": 62756
+    }
   ],
-  "total_files": 1,
-  "total_chars": 62756,
-  "original_path": "books/book_20260922_163508_a82865/original.txt"
+  "raw_original_path": "books/book_20260922_163508_a82865/original.txt",
+  "raw_manifest_path": "books/book_20260922_163508_a82865/manifest.json",
+  "source_map_mode": "initial_offsets",
+  "original_chars": 62756,
+  "prepared_chars": 62756,
+  "analysis": {
+    "chars": 62756,
+    "chars_no_spaces": 29866,
+    "words": 5346,
+    "paragraphs": 232,
+    "sentences": 215,
+    "diacritics": 0
+  },
+  "punctuation_applied": false,
+  "diacritization_applied": false,
+  "characters": [],
+  "dialogue_config": {
+    "enabled": false,
+    "mode": "single_narrator"
+  },
+  "default_voice_settings": {
+    "speaker_id": "",
+    "character_id": "",
+    "voice_id": "",
+    "voice_profile": "",
+    "language": "",
+    "style": "",
+    "speed": 1.0,
+    "pitch": 0.0,
+    "volume": 1.0
+  },
+  "emotion_config": {
+    "enabled": false,
+    "default_emotion": "neutral",
+    "delivery": ""
+  },
+  "correction_config": {
+    "enabled": true,
+    "preserve_revisions": true
+  },
+  "status": "PREPARED",
+  "manifest_updated_at": "2026-09-22T16:35:10.134855+00:00"
 }

books/book_20260922_163508_a82865/prepared/full.txt

@@ -0,0 +1,1124 @@
+The Complete Guide
+to Building Skills
+for Claude
+Contents
+Introduction                    3
+
+Fundamentals                   4
+
+Planning and design             7
+
+Testing and iteration          14
+
+Distribution and sharing       18
+
+Patterns and troubleshooting   21
+
+Resources and references       28
+
+
+
+
+                                    2
+Introduction
+A skill is a set of instructions - packaged as a simple folder - that teaches Claude   Two Paths Through This Guide
+how to handle specific tasks or workflows. Skills are one of the most powerful         Building standalone skills? Focus on Fundamentals, Planning and Design, and
+ways to customize Claude for your specific needs. Instead of re-explaining your        category 1-2. Enhancing an MCP integration? The "Skills + MCP" section and
+preferences, processes, and domain expertise in every conversation, skills let you     category 3 are for you. Both paths share the same technical requirements, but
+teach Claude once and benefit every time.                                              you choose what's relevant to your use case.
+
+Skills are powerful when you have repeatable workflows: generating frontend            What you'll get out of this guide: By the end, you'll be able to build a functional
+designs from specs, conducting research with consistent methodology, creating          skill in a single sitting. Expect about 15-30 minutes to build and test your first
+documents that follow your team's style guide, or orchestrating multi-step             working skill using the skill-creator.
+processes. They work well with Claude's built-in capabilities like code execution
+and document creation. For those building MCP integrations, skills add another         Let's get started.
+
+powerful layer helping turn raw tool access into reliable, optimized workflows.
+
+This guide covers everything you need to know to build effective skills - from
+planning and structure to testing and distribution. Whether you're building a
+skill for yourself, your team, or for the community, you'll find practical patterns
+and real-world examples throughout.
+
+What you'll learn:
+
+• Technical requirements and best practices for skill structure
+
+• Patterns for standalone skills and MCP-enhanced workflows
+
+• Patterns we've seen work well across different use cases
+
+• How to test, iterate, and distribute your skills
+
+Who this is for:
+
+• Developers who want Claude to follow specific workflows consistently
+
+• Power users who want Claude to follow specific workflows
+
+• Teams looking to standardize how Claude works across their organization
+Chapter 1
+
+
+
+
+Fundamentals   4
+  Chapter 1
+
+
+
+Fundamentals
+What is a skill?                                                                    Composability
+
+A skill is a folder containing:                                                     Claude can load multiple skills simultaneously. Your skill should work well
+                                                                                    alongside others, not assume it's the only capability available.
+• SKILL.md (required): Instructions in Markdown with YAML frontmatter
+
+• scripts/ (optional): Executable code (Python, Bash, etc.)                         Portability
+• references/ (optional): Documentation loaded as needed
+                                                                                    Skills work identically across Claude.ai, Claude Code, and API. Create a skill once
+• assets/ (optional): Templates, fonts, icons used in output                        and it works across all surfaces without modification, provided the environment
+                                                                                    supports any dependencies the skill requires.
+
+Core design principles
+                                                                                    For MCP Builders: Skills + Connectors
+Progressive Disclosure
+                                                                                    💡 Building standalone skills without MCP? Skip to Planning and Design - you can
+Skills use a three-level system:                                                    always return here later.
+• First level (YAML frontmatter): Always loaded in Claude's system prompt.
+                                                                                    If you already have a working MCP server, you've done the hard part. Skills are
+  Provides just enough information for Claude to know when each skill should
+                                                                                    the knowledge layer on top - capturing the workflows and best practices you
+  be used without loading all of it into context.
+                                                                                    already know, so Claude can apply them consistently.
+• Second level (SKILL.md body): Loaded when Claude thinks the skill is
+  relevant to the current task. Contains the full instructions and guidance.
+                                                                                    The kitchen analogy
+• Third level (Linked files): Additional files bundled within the skill directory
+  that Claude can choose to navigate and discover only as needed.                   MCP provides the professional kitchen: access to tools, ingredients, and
+                                                                                    equipment.
+This progressive disclosure minimizes token usage while maintaining
+specialized expertise.                                                              Skills provide the recipes: step-by-step instructions on how to create something
+                                                                                    valuable.
+
+
+
+
+                                                                                                                                                                      5
+Together, they enable users to accomplish complex tasks without needing to
+figure out every step themselves.
+
+How they work together:
+
+
+  MCP (Connectivity)                        Skills (Knowledge)
+
+
+  Connects Claude to your service           Teaches Claude how to use your service
+  (Notion, Asana, Linear, etc.)             effectively
+
+
+  Provides real-time data access and tool   Captures workflows and best practices
+  invocation
+
+
+  What Claude can do                        How Claude should do it
+
+
+
+
+Why this matters for your MCP users
+
+Without skills:
+• Users connect your MCP but don't know what to do next
+
+• Support tickets asking "how do I do X with your integration"
+
+• Each conversation starts from scratch
+
+• Inconsistent results because users prompt differently each time
+
+• Users blame your connector when the real issue is workflow guidance
+
+
+With skills:
+• Pre-built workflows activate automatically when needed
+
+• Consistent, reliable tool usage
+
+• Best practices embedded in every interaction
+
+• Lower learning curve for your integration
+
+
+
+
+                                                                                     6
+Chapter 2
+
+
+
+
+Planning and design   7
+  Chapter 2
+
+
+
+Planning and design
+Start with use cases                                                                 Common skill use case categories
+Before writing any code, identify 2-3 concrete use cases your skill should enable.   At Anthropic, we’ve observed three common use cases:
+
+
+Good use case definition:                                                            Category 1: Document & Asset Creation
+
+                                                                                     Used for: Creating consistent, high-quality output including documents,
+   Use Case: Project Sprint Planning                                                 presentations, apps, designs, code, etc.
+   Trigger: User says "help me plan this sprint" or "create
+   sprint tasks"                                                                     Real example: frontend-design skill (also see skills for docx, pptx, xlsx, and
+   Steps:                                                                            ppt)
+   1. Fetch current project status from Linear (via MCP)
+   2. Analyze team velocity and capacity                                             "Create distinctive, production-grade frontend interfaces with high design
+   3. Suggest task prioritization                                                    quality. Use when building web components, pages, artifacts, posters, or
+   4. Create tasks in Linear with proper labels and estimates                        applications."
+   Result: Fully planned sprint with tasks created
+                                                                                     Key techniques:
+
+                                                                                     • Embedded style guides and brand standards
+Ask yourself:                                                                        • Template structures for consistent output
+• What does a user want to accomplish?
+                                                                                     • Quality checklists before finalizing
+• What multi-step workflows does this require?
+                                                                                     • No external tools required - uses Claude's built-in capabilities
+• Which tools are needed (built-in or MCP?)
+
+• What domain knowledge or best practices should be embedded?
+
+
+
+
+                                                                                                                                                                      8
+Category 2: Workflow Automation                                                  Define success criteria
+Used for: Multi-step processes that benefit from consistent methodology,
+                                                                                 How will you know your skill is working?
+including coordination across multiple MCP servers.
+                                                                                 These are aspirational targets - rough benchmarks rather than precise
+Real example: skill-creator skill
+                                                                                 thresholds. Aim for rigor but accept that there will be an element of vibes-based
+
+"Interactive guide for creating new skills. Walks the user through use case      assessment. We are actively developing more robust measurement guidance and
+
+definition, frontmatter generation, instruction writing, and validation."        tooling.
+
+
+Key techniques:
+                                                                                 Quantitative metrics:
+                                                                                 • Skill triggers on 90% of relevant queries
+• Step-by-step workflow with validation gates
+                                                                                   – How to measure: Run 10-20 test queries that should trigger your skill. Track
+• Templates for common structures                                                    how many times it loads automatically vs. requires explicit invocation.
+• Built-in review and improvement suggestions                                    • Completes workflow in X tool calls
+• Iterative refinement loops                                                       – How to measure: Compare the same task with and without the skill enabled.
+                                                                                     Count tool calls and total tokens consumed.
+Category 3: MCP Enhancement                                                      • 0 failed API calls per workflow
+                                                                                   – How to measure: Monitor MCP server logs during test runs. Track retry rates
+Used for: Workflow guidance to enhance the tool access an MCP server provides.
+                                                                                     and error codes.
+Real example: sentry-code-review skill (from Sentry)
+                                                                                 Qualitative metrics:
+"Automatically analyzes and fixes detected bugs in GitHub Pull Requests using
+                                                                                 • Users don't need to prompt Claude about next steps
+Sentry's error monitoring data via their MCP server."
+                                                                                   – How to assess: During testing, note how often you need to redirect or clarify.
+                                                                                     Ask beta users for feedback.
+Key techniques:
+
+• Coordinates multiple MCP calls in sequence
+                                                                                 • Workflows complete without user correction
+• Embeds domain expertise                                                          – How to assess: Run the same request 3-5 times. Compare outputs for
+
+• Provides context users would otherwise need to specify                             structural consistency and quality.
+
+• Error handling for common MCP issues                                           • Consistent results across sessions
+                                                                                   – How to assess: Can a new user accomplish the task on first try with minimal
+                                                                                     guidance?
+
+
+
+
+                                                                                                                                                                      9
+Technical requirements                                                            YAML frontmatter: The most important part
+
+                                                                                  The YAML frontmatter is how Claude decides whether to load your skill. Get this
+File structure
+                                                                                  right.
+
+   your-skill-name/                                                               Minimal required format
+   ├── SKILL.md                  # Required - main skill file
+   ├── scripts/                  # Optional - executable code
+                                                                                     ---
+   │    ├── process_data.py # Example
+                                                                                     name: your-skill-name
+   │    └── validate.sh # Example
+                                                                                     description: What it does. Use when user asks to [specific
+   ├── references/               # Optional - documentation
+                                                                                     phrases].
+   │    ├── api-guide.md # Example
+                                                                                     ---
+   │    └── examples/ # Example
+   └── assets/                  # Optional - templates, etc.
+        └── report-template.md # Example
+                                                                                  That's all you need to start.
+
+
+Critical rules                                                                    Field requirements
+
+                                                                                  name (required):
+SKILL.md naming:
+                                                                                  • kebab-case only
+• Must be exactly SKILL.md (case-sensitive)
+                                                                                  • No spaces or capitals
+• No variations accepted (SKILL.MD, skill.md, etc.)
+                                                                                  • Should match folder name
+
+Skill folder naming:
+                                                                                  description (required):
+• Use kebab-case: notion-project-setup        ✅                                   • MUST include BOTH:
+• No spaces: Notion Project Setup       ❌                                           – What the skill does
+
+• No underscores: notion_project_setup        ❌                                     – When to use it (trigger conditions)
+
+• No capitals: NotionProjectSetup      ❌                                          • Under 1024 characters
+
+                                                                                  • No XML tags (< or >)
+No README.md:                                                                     • Include specific tasks users might say
+• Don't include README.md inside your skill folder                                • Mention file types if relevant
+• All documentation goes in SKILL.md or references/
+
+• Note: when distributing via GitHub, you'll still want a repo-level README for
+  human users — see Distribution and Sharing.
+
+                                                                                                                                                               10
+license (optional):                                                            Writing effective skills
+• Use if making skill open source
+                                                                               The description field
+• Common: MIT, Apache-2.0
+                                                                               According to Anthropic's engineering blog: "This metadata...provides just
+compatibility (optional)
+                                                                               enough information for Claude to know when each skill should be used without
+• 1-500 characters                                                             loading all of it into context." This is the first level of progressive disclosure.
+• Indicates environment requirements: e.g. intended product, required system
+  packages, network access needs, etc.                                         Structure:
+
+metadata (optional):
+                                                                                  [What it does] + [When to use it] + [Key capabilities]
+• Any custom key-value pairs
+
+• Suggested: author, version, mcp-server
+
+• Example:
+                                                                               Examples of good descriptions:
+  ```yaml
+  metadata:
+         author: ProjectHub
+                                                                                  # Good - specific and actionable
+         version: 1.0.0 mcp-server: projecthub                                    description: Analyzes Figma design files and generates
+  ```                                                                             developer handoff documentation. Use when user uploads .fig
+                                                                                  files, asks for "design specs", "component documentation", or
+Security restrictions                                                             "design-to-code handoff".
+
+
+Forbidden in frontmatter:                                                         # Good - includes trigger phrases
+                                                                                  description: Manages Linear project workflows including sprint
+• XML angle brackets (< >)
+                                                                                  planning, task creation, and status tracking. Use when user
+• Skills with "claude" or "anthropic" in name (reserved)                          mentions "sprint", "Linear tasks", "project planning", or asks
+                                                                                  to "create tickets".
+Why: Frontmatter appears in Claude's system prompt. Malicious content could
+inject instructions.                                                              # Good - clear value proposition
+                                                                                  description: End-to-end customer onboarding workflow for
+                                                                                  PayFlow. Handles account creation, payment setup, and
+                                                                                  subscription management. Use when user says "onboard new
+                                                                                  customer", "set up subscription", or "create PayFlow account".
+
+
+
+
+                                                                                                                                                                     11
+Examples of bad descriptions:
+                                                                                       Example:
+                                                                                       ```bash
+   # Too vague
+                                                                                       python scripts/fetch_data.py --project-id PROJECT_ID
+   description: Helps with projects.
+                                                                                       Expected output: [describe what success looks like]
+
+   # Missing triggers
+   description: Creates sophisticated multi-page documentation
+   systems.                                                                         (Add more steps as needed)
+
+
+   # Too technical, no user triggers
+                                                                                    Examples
+   description: Implements the Project entity model with
+   hierarchical relationships.
+                                                                                    Example 1: [common scenario]
+
+                                                                                    User says: "Set up a new marketing campaign"
+
+Writing the main instructions                                                       Actions:
+
+After the frontmatter, write the actual instructions in Markdown.                   1. Fetch existing campaigns via MCP
+
+                                                                                    2. Create new campaign with provided parameters
+Recommended structure:
+                                                                                    Result: Campaign created with confirmation link
+Adapt this template for your skill. Replace bracketed sections with your specific
+content.                                                                            (Add more examples as needed)
+
+
+   ---
+                                                                                    Troubleshooting
+   name: your-skill
+   description: [...]
+   ---
+                                                                                    Error: [Common error message]
+
+                                                                                    Cause: [Why it happens]
+   # Your Skill Name
+
+                                                                                    Solution: [How to fix]
+   ## Instructions
+                                                                                    (Add more error cases as needed)
+   ### Step 1: [First Major Step]
+   Clear explanation of what happens.
+
+
+
+
+                                                                                                                                              12
+Best Practices for Instructions                                    Reference bundled resources clearly
+
+
+Be Specific and Actionable
+                                                                      Before writing queries, consult `references/api-patterns.md`
+                                                                      for:
+✅ Good:                                                               - Rate limiting guidance
+                                                                      - Pagination patterns
+   Run `python scripts/validate.py --input {filename}` to check       - Error codes and handling
+   data format.
+   If validation fails, common issues include:
+   - Missing required fields (add them to the CSV)                 Use progressive disclosure
+   - Invalid date formats (use YYYY-MM-DD)
+                                                                   Keep SKILL.md focused on core instructions. Move detailed documentation to
+
+❌ Bad:                                                             `references/` and link to it. (See Core Design Principles for how the three-
+                                                                   level system works.)
+
+
+   Validate the data before proceeding.
+
+
+
+Include error handling
+
+
+
+   ## Common Issues
+
+
+   ### MCP Connection Failed
+   If you see "Connection refused":
+   1. Verify MCP server is running: Check Settings > Extensions
+   2. Confirm API key is valid
+   3. Try reconnecting: Settings > Extensions > [Your Service] >
+   Reconnect
+
+
+
+
+                                                                                                                                                  13
+Chapter 3
+
+
+
+
+Testing and iteration   14
+  Chapter 3
+
+
+
+Testing and iteration
+Skills can be tested at varying levels of rigor depending on your needs:             Recommended Testing Approach
+• Manual testing in Claude.ai - Run queries directly and observe behavior. Fast
+                                                                                     Based on early experience, effective skills testing typically covers three areas:
+  iteration, no setup required.
+
+• Scripted testing in Claude Code - Automate test cases for repeatable               1. Triggering tests
+  validation across changes.
+                                                                                     Goal: Ensure your skill loads at the right times.
+• Programmatic testing via skills API - Build evaluation suites that run
+  systematically against defined test sets.                                          Test cases:
+
+Choose the approach that matches your quality requirements and the visibility        • ✅ Triggers on obvious tasks
+of your skill. A skill used internally by a small team has different testing needs   • ✅ Triggers on paraphrased requests
+
+                                                                                     • ❌ Doesn't trigger on unrelated topics
+than one deployed to thousands of enterprise users.
+
+
+                                                                                     Example test suite:
+   Pro Tip: Iterate on a single task before expanding
+
+                                                                                         Should trigger:
+We’ve found that the most effective skill creators iterate on a single challenging       - "Help me set up a new ProjectHub workspace"
+task until Claude succeeds, then extract the winning approach into a skill. This         - "I need to create a project in ProjectHub"
+leverages Claude’s in-context learning and provides faster signal than broad             - "Initialize a ProjectHub project for Q4 planning"
+
+testing. Once you have a working foundation, expand to multiple test cases for
+                                                                                         Should NOT trigger:
+coverage.
+                                                                                         - "What's the weather in San Francisco?"
+                                                                                         - "Help me write Python code"
+                                                                                         - "Create a spreadsheet" (unless ProjectHub skill handles
+                                                                                         sheets)
+
+
+
+
+                                                                                                                                                                         15
+2. Functional tests
+                                                                                  With skill:
+Goal: Verify the skill produces correct outputs.                                  - Automatic workflow execution
+                                                                                  - 2 clarifying questions only
+Test cases:                                                                       - 0 failed API calls
+
+• Valid outputs generated                                                         - 6,000 tokens consumed
+
+• API calls succeed
+
+• Error handling works
+
+• Edge cases covered                                                           Using the skill-creator skill
+
+Example:                                                                       The skill-creator skill - available in Claude.ai via plugin directory or
+                                                                               download for Claude Code - can help you build and iterate on skills. If you
+                                                                               have an MCP server and know your top 2–3 workflows, you can build and test a
+   Test: Create project with 5 tasks                                           functional skill in a single sitting - often in 15–30 minutes.
+   Given: Project name "Q4 Planning", 5 task descriptions
+   When: Skill executes workflow                                               Creating skills:
+   Then:
+                                                                               • Generate skills from natural language descriptions
+      - Project created in ProjectHub
+      - 5 tasks created with correct properties                                • Produce properly formatted SKILL.md with frontmatter
+      - All tasks linked to project
+                                                                               • Suggest trigger phrases and structure
+      - No API errors
+
+                                                                               Reviewing skills:
+
+                                                                               • Flag common issues (vague descriptions, missing triggers, structural
+3. Performance comparison                                                        problems)
+
+                                                                               • Identify potential over/under-triggering risks
+Goal: Prove the skill improves results vs. baseline.
+                                                                               • Suggest test cases based on the skill's stated purpose
+Use the metrics from Define Success Criteria. Here's what a comparison might
+look like.                                                                     Iterative improvement:
+
+                                                                               • After using your skill and encountering edge cases or failures, bring those
+Baseline comparison:
+                                                                                 examples back to skill-creator
+
+                                                                               • Example: "Use the issues & solution identified in this chat to improve how the
+   Without skill:
+   - User provides instructions each time
+                                                                                 skill handles [specific edge case]"
+   - 15 back-and-forth messages
+   - 3 failed API calls requiring retry
+   - 12,000 tokens consumed
+
+                                                                                                                                                                  16
+To use:                                                                         Execution issues:
+
+                                                                                • Inconsistent results
+   "Use the skill-creator skill to help me build a skill for
+                                                                                • API call failures
+   [your use case]"
+                                                                                • User corrections needed
+
+
+Note: skill-creator helps you design and refine skills but does not execute
+                                                                                   Solution: Improve instructions, add error handling
+automated test suites or produce quantitative evaluation results.
+
+
+Iteration based on feedback
+Skills are living documents. Plan to iterate based on:
+
+
+Undertriggering signals:
+
+• Skill doesn't load when it should
+
+• Users manually enabling it
+
+• Support questions about when to use it
+
+
+
+   Solution: Add more detail and nuance to the description - this may include
+   keywords particularly for technical terms
+
+
+
+Overtriggering signals:
+
+• Skill loads for irrelevant queries
+
+• Users disabling it
+
+• Confusion about purpose
+
+
+
+   Solution: Add negative triggers, be more specific
+
+
+
+
+                                                                                                                                        17
+Chapter 4
+
+
+
+
+Distribution and
+sharing            18
+  Chapter 4
+
+
+
+Distribution and sharing
+Skills make your MCP integration more complete. As users compare connectors,         Using skills via API
+those with skills offer a faster path to value, giving you an edge over MCP-only
+alternatives.                                                                        For programmatic use cases - such as building applications, agents, or automated
+                                                                                     workflows that leverage skills - the API provides direct control over skill
+                                                                                     management and execution.
+Current distribution model (January 2026)
+                                                                                     Key capabilities:
+How individual users get skills:
+                                                                                     •    `/v1/skills` endpoint for listing and managing skills
+1. Download the skill folder
+                                                                                     • Add skills to Messages API requests via the `container.skills` parameter
+2. Zip the folder (if needed)
+                                                                                     • Version control and management through the Claude Console
+3. Upload to Claude.ai via Settings > Capabilities > Skills
+                                                                                     • Works with the Claude Agent SDK for building custom agents
+4. Or place in Claude Code skills directory
+
+                                                                                     When to use skills via the API vs. Claude.ai:
+Organization-level skills:
+• Admins can deploy skills workspace-wide (shipped December 18, 2025)                    Use Case                                           Best Surface
+• Automatic updates
+                                                                                         End users interacting with skills directly         Claude.ai / Claude Code
+• Centralized management
+                                                                                         Manual testing and iteration during development    Claude.ai / Claude Code
+
+An open standard
+                                                                                         Individual, ad-hoc workflows                       Claude.ai / Claude Code
+We've published Agent Skills as an open standard. Like MCP, we believe skills
+should be portable across tools and platforms - the same skill should work               Applications using skills programmatically         API
+whether you're using Claude or other AI platforms. That said, some skills are
+designed to take full advantage of a specific platform's capabilities; authors can       Production deployments at scale                    API
+
+note this in the skill's compatibility field. We've been collaborating with
+members of the ecosystem on the standard, and we're excited by early adoption.           Automated pipelines and agent systems              API
+
+
+
+
+                                                                                                                                                                      19
+Note: Skills in the API require the Code Execution Tool beta, which provides the
+secure environment skills need to run.                                                      - Select the skill folder (zipped)
+
+
+For implementation details, see:                                                        3. Enable the skill:
+                                                                                            - Toggle on the [Your Service] skill
+• Skills API Quickstart                                                                     - Ensure your MCP server is connected
+• Create Custom skills
+                                                                                        4. Test:
+• Skills in the Agent SDK                                                                   - Ask Claude: "Set up a new project in [Your Service]"
+
+
+
+Recommended approach today                                                           Positioning your skill
+Start by hosting your skill on GitHub with a public repo, clear README (for          How you describe your skill determines whether users understand its value and
+human visitors — this is separate from your skill folder, which should not contain   actually try it. When writing about your skill—in your README, documentation,
+a README.md), and example usage with screenshots. Then add a section                 or marketing - keep these principles in mind.
+to your MCP documentation that links to the skill, explains why using both
+together is valuable, and provides a quick-start guide.                              Focus on outcomes, not features:
+
+1. Host on GitHub                                                                    ✅ Good:
+  – Public repo for open-source skills
+  – Clear README with installation instructions                                         "The ProjectHub skill enables teams to set up complete project
+  – Example usage and screenshots                                                       workspaces in seconds — including pages, databases, and
+                                                                                        templates — instead of spending 30 minutes on manual setup."
+2. Document in Your MCP Repo
+  – Link to skills from MCP documentation
+  – Explain the value of using both together
+  – Provide quick-start guide
+                                                                                     ❌ Bad:
+3. Create an Installation Guide                                                         "The ProjectHub skill is a folder containing YAML frontmatter
+                                                                                        and Markdown instructions that calls our MCP server tools."
+
+   ## Installing the [Your Service] skill
+
+   1. Download the skill:                                                            Highlight the MCP + skills story:
+       - Clone repo: `git clone https://github.com/yourcompany/
+         skills`
+       - Or download ZIP from Releases                                                  "Our MCP server gives Claude access to your Linear projects.
+                                                                                        Our skills teach Claude your team's sprint planning workflow.
+   2. Install in Claude:                                                                Together, they enable AI-powered project management."
+       - Open Claude.ai > Settings > skills
+       - Click "Upload skill"
+                                                                                                                                                                 20
+Chapter 5
+
+
+
+
+Patterns and
+troubleshooting   21
+  Chapter 5
+
+
+
+Patterns and troubleshooting
+These patterns emerged from skills created by early adopters and internal teams.    Pattern 1: Sequential workflow orchestration
+They represent common approaches we've seen work well, not prescriptive
+templates.                                                                          Use when: Your users need multi-step processes in a specific order.
+
+                                                                                    Example structure:
+Choosing your approach: Problem-first vs. tool-first
+
+Think of it like Home Depot. You might walk in with a problem - "I need to fix a       ## Workflow: Onboard New Customer
+kitchen cabinet" - and an employee points you to the right tools. Or you might
+pick out a new drill and ask how to use it for your specific job.                      ### Step 1: Create Account
+                                                                                       Call MCP tool: `create_customer`
+Skills work the same way:                                                              Parameters: name, email, company
+
+• Problem-first: "I need to set up a project workspace" → Your skill orchestrates
+                                                                                       ### Step 2: Setup Payment
+  the right MCP calls in the right sequence. Users describe outcomes; the skill
+                                                                                       Call MCP tool: `setup_payment_method`
+  handles the tools.
+                                                                                       Wait for: payment method verification
+• Tool-first: "I have Notion MCP connected" → Your skill teaches Claude the
+  optimal workflows and best practices. Users have access; the skill provides          ### Step 3: Create Subscription
+  expertise.                                                                           Call MCP tool: `create_subscription`
+                                                                                       Parameters: plan_id, customer_id (from Step 1)
+Most skills lean one direction. Knowing which framing fits your use case helps
+you choose the right pattern below.                                                    ### Step 4: Send Welcome Email
+                                                                                       Call MCP tool: `send_email`
+                                                                                       Template: welcome_email_template
+
+
+
+
+                                                                                    Key techniques:
+
+                                                                                    • Explicit step ordering
+
+                                                                                    • Dependencies between steps
+
+                                                                                    • Validation at each stage
+
+                                                                                    • Rollback instructions for failures
+                                                                                                                                                          22
+Pattern 2: Multi-MCP coordination               Pattern 3: Iterative refinement
+Use when: Workflows span multiple services.     Use when: Output quality improves with iteration.
+
+Example: Design-to-development handoff          Example: Report generation
+
+
+   ### Phase 1: Design Export (Figma MCP)          ## Iterative Report Creation
+   1. Export design assets from Figma
+   2. Generate design specifications               ### Initial Draft
+   3. Create asset manifest                        1. Fetch data via MCP
+                                                   2. Generate first draft report
+   ### Phase 2: Asset Storage (Drive MCP)          3. Save to temporary file
+   1. Create project folder in Drive
+   2. Upload all assets                            ### Quality Check
+   3. Generate shareable links                     1. Run validation script: `scripts/check_report.py`
+                                                   2. Identify issues:
+   ### Phase 3: Task Creation (Linear MCP)             - Missing sections
+   1. Create development tasks                         - Inconsistent formatting
+   2. Attach asset links to tasks                      - Data validation errors
+   3. Assign to engineering team
+                                                   ### Refinement Loop
+   ### Phase 4: Notification (Slack MCP)           1. Address each identified issue
+   1. Post handoff summary to #engineering         2. Regenerate affected sections
+   2. Include asset links and task references      3. Re-validate
+                                                   4. Repeat until quality threshold met
+
+
+Key techniques:                                    ### Finalization
+                                                   1. Apply final formatting
+• Clear phase separation
+                                                   2. Generate summary
+• Data passing between MCPs                        3. Save final version
+
+• Validation before moving to next phase
+
+• Centralized error handling                    Key techniques:
+
+                                                • Explicit quality criteria
+
+                                                • Iterative improvement
+
+                                                • Validation scripts
+
+                                                • Know when to stop iterating
+
+                                                                                                         23
+Pattern 4: Context-aware tool selection                         Pattern 5: Domain-specific intelligence
+Use when: Same outcome, different tools depending on context.   Use when: Your skill adds specialized knowledge beyond tool access.
+
+Example: File storage                                           Example: Financial compliance
+
+
+   ## Smart File Storage                                           ## Payment Processing with Compliance
+
+
+   ### Decision Tree                                               ### Before Processing (Compliance Check)
+   1. Check file type and size                                     1. Fetch transaction details via MCP
+   2. Determine best storage location:                             2. Apply compliance rules:
+       - Large files (>10MB): Use cloud storage MCP                   - Check sanctions lists
+       - Collaborative docs: Use Notion/Docs MCP                      - Verify jurisdiction allowances
+       - Code files: Use GitHub MCP                                   - Assess risk level
+       - Temporary files: Use local storage                        3. Document compliance decision
+
+
+   ### Execute Storage                                             ### Processing
+   Based on decision:                                              IF compliance passed:
+   - Call appropriate MCP tool                                       - Call payment processing MCP tool
+   - Apply service-specific metadata                                 - Apply appropriate fraud checks
+   - Generate access link                                            - Process transaction
+                                                                   ELSE:
+   ### Provide Context to User                                       - Flag for review
+   Explain why that storage was chosen                               - Create compliance case
+
+
+                                                                   ### Audit Trail
+Key techniques:                                                    - Log all compliance checks
+                                                                   - Record processing decisions
+• Clear decision criteria
+                                                                   - Generate audit report
+• Fallback options
+
+• Transparency about choices
+                                                                Key techniques:
+
+                                                                • Domain expertise embedded in logic
+
+                                                                • Compliance before action
+
+                                                                • Comprehensive documentation
+
+                                                                • Clear governance
+
+
+                                                                                                                                      24
+Troubleshooting
+                                                         # Wrong
+                                                         name: My Cool Skill
+Skill won't upload
+
+Error: "Could not find SKILL.md in uploaded folder"      # Correct
+                                                         name: my-cool-skill
+Cause: File not named exactly SKILL.md
+
+Solution:
+                                                      Skill doesn't trigger
+• Rename to SKILL.md (case-sensitive)
+
+• Verify with: ls -la should show SKILL.md            Symptom: Skill never loads automatically
+
+Error: "Invalid frontmatter"                          Fix:
+
+Cause: YAML formatting issue                          Revise your description field. See The Description Field for good/bad examples.
+
+Common mistakes:                                      Quick checklist:
+
+                                                      • Is it too generic? ("Helps with projects" won't work)
+   # Wrong - missing delimiters                       • Does it include trigger phrases users would actually say?
+   name: my-skill
+   description: Does things
+                                                      • Does it mention relevant file types if applicable?
+
+                                                      Debugging approach:
+   # Wrong - unclosed quotes
+   name: my-skill
+                                                      Ask Claude: "When would you use the [skill name] skill?" Claude will quote the
+   description: "Does things
+                                                      description back. Adjust based on what's missing.
+
+   # Correct
+   ---
+                                                      Skill triggers too often
+   name: my-skill
+                                                      Symptom: Skill loads for unrelated queries
+   description: Does things
+   ---                                                Solutions:
+
+                                                      1. Add negative triggers
+Error: "Invalid skill name"
+                                                         description: Advanced data analysis for CSV files. Use for
+Cause: Name has spaces or capitals                       statistical modeling, regression, clustering. Do NOT use for
+                                                         simple data exploration (use data-viz skill instead).
+
+
+
+                                                                                                                                       25
+2. Be more specific                                                 Instructions not followed
+
+                                                                    Symptom: Skill loads but Claude doesn't follow instructions
+   # Too broad
+   description: Processes documents
+                                                                    Common causes:
+
+   # More specific                                                  1. Instructions too verbose
+   description: Processes PDF legal documents for contract review     – Keep instructions concise
+                                                                      – Use bullet points and numbered lists
+                                                                      – Move detailed reference to separate files
+3. Clarify scope
+                                                                    2. Instructions buried
+                                                                      – Put critical instructions at the top
+   description: PayFlow payment processing for e-commerce. Use        – Use ## Important or ## Critical headers
+   specifically for online payment workflows, not for general
+                                                                      – Repeat key points if needed
+   financial queries.
+                                                                    3. Ambiguous language
+
+
+                                                                       # Bad
+MCP connection issues                                                  Make sure to validate things properly
+
+Symptom: Skill loads but MCP calls fail
+                                                                       # Good
+                                                                       CRITICAL: Before calling create_project, verify:
+Checklist:
+                                                                       - Project name is non-empty
+1. Verify MCP server is connected                                      - At least one team member assigned
+  – Claude.ai: Settings > Extensions > [Your Service]                  - Start date is not in the past
+  – Should show "Connected" status
+
+2. Check authentication                                             Advanced technique: For critical validations, consider bundling a script
+  – API keys valid and not expired                                  that performs the checks programmatically rather than relying on language
+  – Proper permissions/scopes granted                               instructions. Code is deterministic; language interpretation isn't. See the Office
+  – OAuth tokens refreshed                                          skills for examples of this pattern.
+
+3. Test MCP independently                                           4. Model "laziness" Add explicit encouragement:
+  – Ask Claude to call MCP directly (without skill)
+  – "Use [Service] MCP to fetch my projects"                           ## Performance Notes
+  – If this fails, issue is MCP not skill                              - Take your time to do this thoroughly
+                                                                       - Quality is more important than speed
+4. Verify tool names
+                                                                       - Do not skip validation steps
+  – Skill references correct MCP tool names
+  – Check MCP server documentation
+                                                                    Note: Adding this to user prompts is more effective than in SKILL.md
+  – Tool names are case-sensitive
+                                                                                                                                                         26
+Large context issues
+
+Symptom: Skill seems slow or responses degraded
+
+Causes:
+
+• Skill content too large
+
+• Too many skills enabled simultaneously
+
+• All content loaded instead of progressive disclosure
+
+Solutions:
+
+1. Optimize SKILL.md size
+  – Move detailed docs to references/
+  – Link to references instead of inline
+  – Keep SKILL.md under 5,000 words
+
+2. Reduce enabled skills
+  – Evaluate if you have more than 20 - 50 skills enabled simultaneously
+  – Recommend selective enablement
+  – Consider skill "packs" for related capabilities
+
+
+
+
+                                                                           27
+Chapter 6
+
+
+
+
+Resources and
+references      28
+  Chapter 6
+
+
+
+Resources and references
+If you're building your first skill, start with the Best Practices Guide, then   Tools and Utilities
+reference the API docs as needed.
+                                                                                 skill-creator skill:
+Official Documentation                                                           • Built into Claude.ai and available for Claude Code
+
+                                                                                 • Can generate skills from descriptions
+Anthropic Resources:
+                                                                                 • Reviews and provides recommendations
+• Best Practices Guide
+                                                                                 • Use: "Help me build a skill using skill-creator"
+• Skills Documentation
+
+• API Reference                                                                  Validation:
+• MCP Documentation                                                              • skill-creator can assess your skills
+
+                                                                                 • Ask: "Review this skill and suggest improvements"
+Blog Posts:
+• Introducing Agent Skills
+                                                                                 Getting Support
+• Engineering Blog: Equipping Agents for the Real World
+
+• Skills Explained                                                               For Technical Questions:
+
+• How to Create Skills for Claude                                                • General questions: Community forums at the Claude Developers Discord
+
+• Building Skills for Claude Code
+                                                                                 For Bug Reports:
+• Improving Frontend Design through Skills
+                                                                                 • GitHub Issues: anthropics/skills/issues
+
+                                                                                 • Include: Skill name, error message, steps to reproduce
+Example skills
+
+Public skills repository:
+• GitHub: anthropics/skills
+
+• Contains Anthropic-created skills you can customize
+
+                                                                                                                                                          29
+                                                                                     Before upload
+Reference A: Quick                                                                     Tested triggering on obvious tasks
+
+
+checklist                                                                              Tested triggering on paraphrased requests
+
+                                                                                       Verified doesn't trigger on unrelated topics
+Use this checklist to validate your skill before and after upload. If you want
+                                                                                       Functional tests pass
+a faster start, use the skill-creator skill to generate your first draft, then run
+through this list to make sure you haven't missed anything.                            Tool integration works (if applicable)
+
+                                                                                       Compressed as .zip file
+
+Before you start
+                                                                                     After upload
+    Identified 2-3 concrete use cases
+
+    Tools identified (built-in or MCP)                                                 Test in real conversations
+
+    Reviewed this guide and example skills                                             Monitor for under/over-triggering
+
+    Planned folder structure                                                           Collect user feedback
+
+                                                                                       Iterate on description and instructions
+
+During development                                                                     Update version in metadata
+
+    Folder named in kebab-case
+
+    SKILL.md file exists (exact spelling)
+
+    YAML frontmatter has --- delimiters
+
+    name field: kebab-case, no spaces, no capitals
+
+    description includes WHAT and WHEN
+
+    No XML tags (< >) anywhere
+
+    Instructions are clear and actionable
+
+    Error handling included
+
+    Examples provided
+
+    References clearly linked
+
+
+
+
+                                                                                                                                      30
+                                                                     Security notes
+Reference B: YAML
+frontmatter                                                          Allowed:
+                                                                     • Any standard YAML types (strings, numbers, booleans, lists, objects)
+
+                                                                     • Custom metadata fields
+Required fields
+                                                                     • Long descriptions (up to 1024 characters)
+
+  ---
+                                                                     Forbidden:
+  name: skill-name-in-kebab-case
+  description: What it does and when to use it. Include specific     • XML angle brackets (< >) - security restriction
+  trigger phrases.
+                                                                     • Code execution in YAML (uses safe YAML parsing)
+  ---
+                                                                     • Skills named with "claude" or "anthropic" prefix (reserved)
+
+
+
+
+All optional fields
+
+  name: skill-name
+  description: [required description]
+  license: MIT # Optional: License for open-source
+  allowed-tools: "Bash(python:*) Bash(npm:*) WebFetch" # Optional:
+  Restrict tool access
+  metadata: # Optional: Custom fields
+    author: Company Name
+    version: 1.0.0
+    mcp-server: server-name
+    category: productivity
+    tags: [project-management, automation]
+    documentation: https://example.com/docs
+    support: support@example.com
+
+
+
+
+                                                                                                                                              31
+Reference C: Complete skill
+examples
+For full, production-ready skills demonstrating the patterns in this guide:
+
+• Document Skills - PDF, DOCX, PPTX, XLSX creation
+
+• Example Skills - Various workflow patterns
+
+• Partner Skills Directory - View skills from various partners such as Asana,
+  Atlassian, Canva, Figma, Sentry, Zapier, and more
+
+These repositories stay up-to-date and include additional examples beyond
+what's covered here. Clone them, modify them for your use case, and use them as
+templates.
+
+
+
+
+                                                                                  32
+claude.ai

books/book_20260922_163508_a82865/source/original.txt

@@ -0,0 +1,1124 @@
+The Complete Guide
+to Building Skills
+for Claude
+Contents
+Introduction                    3
+
+Fundamentals                   4
+
+Planning and design             7
+
+Testing and iteration          14
+
+Distribution and sharing       18
+
+Patterns and troubleshooting   21
+
+Resources and references       28
+
+
+
+
+                                    2
+Introduction
+A skill is a set of instructions - packaged as a simple folder - that teaches Claude   Two Paths Through This Guide
+how to handle specific tasks or workflows. Skills are one of the most powerful         Building standalone skills? Focus on Fundamentals, Planning and Design, and
+ways to customize Claude for your specific needs. Instead of re-explaining your        category 1-2. Enhancing an MCP integration? The "Skills + MCP" section and
+preferences, processes, and domain expertise in every conversation, skills let you     category 3 are for you. Both paths share the same technical requirements, but
+teach Claude once and benefit every time.                                              you choose what's relevant to your use case.
+
+Skills are powerful when you have repeatable workflows: generating frontend            What you'll get out of this guide: By the end, you'll be able to build a functional
+designs from specs, conducting research with consistent methodology, creating          skill in a single sitting. Expect about 15-30 minutes to build and test your first
+documents that follow your team's style guide, or orchestrating multi-step             working skill using the skill-creator.
+processes. They work well with Claude's built-in capabilities like code execution
+and document creation. For those building MCP integrations, skills add another         Let's get started.
+
+powerful layer helping turn raw tool access into reliable, optimized workflows.
+
+This guide covers everything you need to know to build effective skills - from
+planning and structure to testing and distribution. Whether you're building a
+skill for yourself, your team, or for the community, you'll find practical patterns
+and real-world examples throughout.
+
+What you'll learn:
+
+• Technical requirements and best practices for skill structure
+
+• Patterns for standalone skills and MCP-enhanced workflows
+
+• Patterns we've seen work well across different use cases
+
+• How to test, iterate, and distribute your skills
+
+Who this is for:
+
+• Developers who want Claude to follow specific workflows consistently
+
+• Power users who want Claude to follow specific workflows
+
+• Teams looking to standardize how Claude works across their organization
+Chapter 1
+
+
+
+
+Fundamentals   4
+  Chapter 1
+
+
+
+Fundamentals
+What is a skill?                                                                    Composability
+
+A skill is a folder containing:                                                     Claude can load multiple skills simultaneously. Your skill should work well
+                                                                                    alongside others, not assume it's the only capability available.
+• SKILL.md (required): Instructions in Markdown with YAML frontmatter
+
+• scripts/ (optional): Executable code (Python, Bash, etc.)                         Portability
+• references/ (optional): Documentation loaded as needed
+                                                                                    Skills work identically across Claude.ai, Claude Code, and API. Create a skill once
+• assets/ (optional): Templates, fonts, icons used in output                        and it works across all surfaces without modification, provided the environment
+                                                                                    supports any dependencies the skill requires.
+
+Core design principles
+                                                                                    For MCP Builders: Skills + Connectors
+Progressive Disclosure
+                                                                                    💡 Building standalone skills without MCP? Skip to Planning and Design - you can
+Skills use a three-level system:                                                    always return here later.
+• First level (YAML frontmatter): Always loaded in Claude's system prompt.
+                                                                                    If you already have a working MCP server, you've done the hard part. Skills are
+  Provides just enough information for Claude to know when each skill should
+                                                                                    the knowledge layer on top - capturing the workflows and best practices you
+  be used without loading all of it into context.
+                                                                                    already know, so Claude can apply them consistently.
+• Second level (SKILL.md body): Loaded when Claude thinks the skill is
+  relevant to the current task. Contains the full instructions and guidance.
+                                                                                    The kitchen analogy
+• Third level (Linked files): Additional files bundled within the skill directory
+  that Claude can choose to navigate and discover only as needed.                   MCP provides the professional kitchen: access to tools, ingredients, and
+                                                                                    equipment.
+This progressive disclosure minimizes token usage while maintaining
+specialized expertise.                                                              Skills provide the recipes: step-by-step instructions on how to create something
+                                                                                    valuable.
+
+
+
+
+                                                                                                                                                                      5
+Together, they enable users to accomplish complex tasks without needing to
+figure out every step themselves.
+
+How they work together:
+
+
+  MCP (Connectivity)                        Skills (Knowledge)
+
+
+  Connects Claude to your service           Teaches Claude how to use your service
+  (Notion, Asana, Linear, etc.)             effectively
+
+
+  Provides real-time data access and tool   Captures workflows and best practices
+  invocation
+
+
+  What Claude can do                        How Claude should do it
+
+
+
+
+Why this matters for your MCP users
+
+Without skills:
+• Users connect your MCP but don't know what to do next
+
+• Support tickets asking "how do I do X with your integration"
+
+• Each conversation starts from scratch
+
+• Inconsistent results because users prompt differently each time
+
+• Users blame your connector when the real issue is workflow guidance
+
+
+With skills:
+• Pre-built workflows activate automatically when needed
+
+• Consistent, reliable tool usage
+
+• Best practices embedded in every interaction
+
+• Lower learning curve for your integration
+
+
+
+
+                                                                                     6
+Chapter 2
+
+
+
+
+Planning and design   7
+  Chapter 2
+
+
+
+Planning and design
+Start with use cases                                                                 Common skill use case categories
+Before writing any code, identify 2-3 concrete use cases your skill should enable.   At Anthropic, we’ve observed three common use cases:
+
+
+Good use case definition:                                                            Category 1: Document & Asset Creation
+
+                                                                                     Used for: Creating consistent, high-quality output including documents,
+   Use Case: Project Sprint Planning                                                 presentations, apps, designs, code, etc.
+   Trigger: User says "help me plan this sprint" or "create
+   sprint tasks"                                                                     Real example: frontend-design skill (also see skills for docx, pptx, xlsx, and
+   Steps:                                                                            ppt)
+   1. Fetch current project status from Linear (via MCP)
+   2. Analyze team velocity and capacity                                             "Create distinctive, production-grade frontend interfaces with high design
+   3. Suggest task prioritization                                                    quality. Use when building web components, pages, artifacts, posters, or
+   4. Create tasks in Linear with proper labels and estimates                        applications."
+   Result: Fully planned sprint with tasks created
+                                                                                     Key techniques:
+
+                                                                                     • Embedded style guides and brand standards
+Ask yourself:                                                                        • Template structures for consistent output
+• What does a user want to accomplish?
+                                                                                     • Quality checklists before finalizing
+• What multi-step workflows does this require?
+                                                                                     • No external tools required - uses Claude's built-in capabilities
+• Which tools are needed (built-in or MCP?)
+
+• What domain knowledge or best practices should be embedded?
+
+
+
+
+                                                                                                                                                                      8
+Category 2: Workflow Automation                                                  Define success criteria
+Used for: Multi-step processes that benefit from consistent methodology,
+                                                                                 How will you know your skill is working?
+including coordination across multiple MCP servers.
+                                                                                 These are aspirational targets - rough benchmarks rather than precise
+Real example: skill-creator skill
+                                                                                 thresholds. Aim for rigor but accept that there will be an element of vibes-based
+
+"Interactive guide for creating new skills. Walks the user through use case      assessment. We are actively developing more robust measurement guidance and
+
+definition, frontmatter generation, instruction writing, and validation."        tooling.
+
+
+Key techniques:
+                                                                                 Quantitative metrics:
+                                                                                 • Skill triggers on 90% of relevant queries
+• Step-by-step workflow with validation gates
+                                                                                   – How to measure: Run 10-20 test queries that should trigger your skill. Track
+• Templates for common structures                                                    how many times it loads automatically vs. requires explicit invocation.
+• Built-in review and improvement suggestions                                    • Completes workflow in X tool calls
+• Iterative refinement loops                                                       – How to measure: Compare the same task with and without the skill enabled.
+                                                                                     Count tool calls and total tokens consumed.
+Category 3: MCP Enhancement                                                      • 0 failed API calls per workflow
+                                                                                   – How to measure: Monitor MCP server logs during test runs. Track retry rates
+Used for: Workflow guidance to enhance the tool access an MCP server provides.
+                                                                                     and error codes.
+Real example: sentry-code-review skill (from Sentry)
+                                                                                 Qualitative metrics:
+"Automatically analyzes and fixes detected bugs in GitHub Pull Requests using
+                                                                                 • Users don't need to prompt Claude about next steps
+Sentry's error monitoring data via their MCP server."
+                                                                                   – How to assess: During testing, note how often you need to redirect or clarify.
+                                                                                     Ask beta users for feedback.
+Key techniques:
+
+• Coordinates multiple MCP calls in sequence
+                                                                                 • Workflows complete without user correction
+• Embeds domain expertise                                                          – How to assess: Run the same request 3-5 times. Compare outputs for
+
+• Provides context users would otherwise need to specify                             structural consistency and quality.
+
+• Error handling for common MCP issues                                           • Consistent results across sessions
+                                                                                   – How to assess: Can a new user accomplish the task on first try with minimal
+                                                                                     guidance?
+
+
+
+
+                                                                                                                                                                      9
+Technical requirements                                                            YAML frontmatter: The most important part
+
+                                                                                  The YAML frontmatter is how Claude decides whether to load your skill. Get this
+File structure
+                                                                                  right.
+
+   your-skill-name/                                                               Minimal required format
+   ├── SKILL.md                  # Required - main skill file
+   ├── scripts/                  # Optional - executable code
+                                                                                     ---
+   │    ├── process_data.py # Example
+                                                                                     name: your-skill-name
+   │    └── validate.sh # Example
+                                                                                     description: What it does. Use when user asks to [specific
+   ├── references/               # Optional - documentation
+                                                                                     phrases].
+   │    ├── api-guide.md # Example
+                                                                                     ---
+   │    └── examples/ # Example
+   └── assets/                  # Optional - templates, etc.
+        └── report-template.md # Example
+                                                                                  That's all you need to start.
+
+
+Critical rules                                                                    Field requirements
+
+                                                                                  name (required):
+SKILL.md naming:
+                                                                                  • kebab-case only
+• Must be exactly SKILL.md (case-sensitive)
+                                                                                  • No spaces or capitals
+• No variations accepted (SKILL.MD, skill.md, etc.)
+                                                                                  • Should match folder name
+
+Skill folder naming:
+                                                                                  description (required):
+• Use kebab-case: notion-project-setup        ✅                                   • MUST include BOTH:
+• No spaces: Notion Project Setup       ❌                                           – What the skill does
+
+• No underscores: notion_project_setup        ❌                                     – When to use it (trigger conditions)
+
+• No capitals: NotionProjectSetup      ❌                                          • Under 1024 characters
+
+                                                                                  • No XML tags (< or >)
+No README.md:                                                                     • Include specific tasks users might say
+• Don't include README.md inside your skill folder                                • Mention file types if relevant
+• All documentation goes in SKILL.md or references/
+
+• Note: when distributing via GitHub, you'll still want a repo-level README for
+  human users — see Distribution and Sharing.
+
+                                                                                                                                                               10
+license (optional):                                                            Writing effective skills
+• Use if making skill open source
+                                                                               The description field
+• Common: MIT, Apache-2.0
+                                                                               According to Anthropic's engineering blog: "This metadata...provides just
+compatibility (optional)
+                                                                               enough information for Claude to know when each skill should be used without
+• 1-500 characters                                                             loading all of it into context." This is the first level of progressive disclosure.
+• Indicates environment requirements: e.g. intended product, required system
+  packages, network access needs, etc.                                         Structure:
+
+metadata (optional):
+                                                                                  [What it does] + [When to use it] + [Key capabilities]
+• Any custom key-value pairs
+
+• Suggested: author, version, mcp-server
+
+• Example:
+                                                                               Examples of good descriptions:
+  ```yaml
+  metadata:
+         author: ProjectHub
+                                                                                  # Good - specific and actionable
+         version: 1.0.0 mcp-server: projecthub                                    description: Analyzes Figma design files and generates
+  ```                                                                             developer handoff documentation. Use when user uploads .fig
+                                                                                  files, asks for "design specs", "component documentation", or
+Security restrictions                                                             "design-to-code handoff".
+
+
+Forbidden in frontmatter:                                                         # Good - includes trigger phrases
+                                                                                  description: Manages Linear project workflows including sprint
+• XML angle brackets (< >)
+                                                                                  planning, task creation, and status tracking. Use when user
+• Skills with "claude" or "anthropic" in name (reserved)                          mentions "sprint", "Linear tasks", "project planning", or asks
+                                                                                  to "create tickets".
+Why: Frontmatter appears in Claude's system prompt. Malicious content could
+inject instructions.                                                              # Good - clear value proposition
+                                                                                  description: End-to-end customer onboarding workflow for
+                                                                                  PayFlow. Handles account creation, payment setup, and
+                                                                                  subscription management. Use when user says "onboard new
+                                                                                  customer", "set up subscription", or "create PayFlow account".
+
+
+
+
+                                                                                                                                                                     11
+Examples of bad descriptions:
+                                                                                       Example:
+                                                                                       ```bash
+   # Too vague
+                                                                                       python scripts/fetch_data.py --project-id PROJECT_ID
+   description: Helps with projects.
+                                                                                       Expected output: [describe what success looks like]
+
+   # Missing triggers
+   description: Creates sophisticated multi-page documentation
+   systems.                                                                         (Add more steps as needed)
+
+
+   # Too technical, no user triggers
+                                                                                    Examples
+   description: Implements the Project entity model with
+   hierarchical relationships.
+                                                                                    Example 1: [common scenario]
+
+                                                                                    User says: "Set up a new marketing campaign"
+
+Writing the main instructions                                                       Actions:
+
+After the frontmatter, write the actual instructions in Markdown.                   1. Fetch existing campaigns via MCP
+
+                                                                                    2. Create new campaign with provided parameters
+Recommended structure:
+                                                                                    Result: Campaign created with confirmation link
+Adapt this template for your skill. Replace bracketed sections with your specific
+content.                                                                            (Add more examples as needed)
+
+
+   ---
+                                                                                    Troubleshooting
+   name: your-skill
+   description: [...]
+   ---
+                                                                                    Error: [Common error message]
+
+                                                                                    Cause: [Why it happens]
+   # Your Skill Name
+
+                                                                                    Solution: [How to fix]
+   ## Instructions
+                                                                                    (Add more error cases as needed)
+   ### Step 1: [First Major Step]
+   Clear explanation of what happens.
+
+
+
+
+                                                                                                                                              12
+Best Practices for Instructions                                    Reference bundled resources clearly
+
+
+Be Specific and Actionable
+                                                                      Before writing queries, consult `references/api-patterns.md`
+                                                                      for:
+✅ Good:                                                               - Rate limiting guidance
+                                                                      - Pagination patterns
+   Run `python scripts/validate.py --input {filename}` to check       - Error codes and handling
+   data format.
+   If validation fails, common issues include:
+   - Missing required fields (add them to the CSV)                 Use progressive disclosure
+   - Invalid date formats (use YYYY-MM-DD)
+                                                                   Keep SKILL.md focused on core instructions. Move detailed documentation to
+
+❌ Bad:                                                             `references/` and link to it. (See Core Design Principles for how the three-
+                                                                   level system works.)
+
+
+   Validate the data before proceeding.
+
+
+
+Include error handling
+
+
+
+   ## Common Issues
+
+
+   ### MCP Connection Failed
+   If you see "Connection refused":
+   1. Verify MCP server is running: Check Settings > Extensions
+   2. Confirm API key is valid
+   3. Try reconnecting: Settings > Extensions > [Your Service] >
+   Reconnect
+
+
+
+
+                                                                                                                                                  13
+Chapter 3
+
+
+
+
+Testing and iteration   14
+  Chapter 3
+
+
+
+Testing and iteration
+Skills can be tested at varying levels of rigor depending on your needs:             Recommended Testing Approach
+• Manual testing in Claude.ai - Run queries directly and observe behavior. Fast
+                                                                                     Based on early experience, effective skills testing typically covers three areas:
+  iteration, no setup required.
+
+• Scripted testing in Claude Code - Automate test cases for repeatable               1. Triggering tests
+  validation across changes.
+                                                                                     Goal: Ensure your skill loads at the right times.
+• Programmatic testing via skills API - Build evaluation suites that run
+  systematically against defined test sets.                                          Test cases:
+
+Choose the approach that matches your quality requirements and the visibility        • ✅ Triggers on obvious tasks
+of your skill. A skill used internally by a small team has different testing needs   • ✅ Triggers on paraphrased requests
+
+                                                                                     • ❌ Doesn't trigger on unrelated topics
+than one deployed to thousands of enterprise users.
+
+
+                                                                                     Example test suite:
+   Pro Tip: Iterate on a single task before expanding
+
+                                                                                         Should trigger:
+We’ve found that the most effective skill creators iterate on a single challenging       - "Help me set up a new ProjectHub workspace"
+task until Claude succeeds, then extract the winning approach into a skill. This         - "I need to create a project in ProjectHub"
+leverages Claude’s in-context learning and provides faster signal than broad             - "Initialize a ProjectHub project for Q4 planning"
+
+testing. Once you have a working foundation, expand to multiple test cases for
+                                                                                         Should NOT trigger:
+coverage.
+                                                                                         - "What's the weather in San Francisco?"
+                                                                                         - "Help me write Python code"
+                                                                                         - "Create a spreadsheet" (unless ProjectHub skill handles
+                                                                                         sheets)
+
+
+
+
+                                                                                                                                                                         15
+2. Functional tests
+                                                                                  With skill:
+Goal: Verify the skill produces correct outputs.                                  - Automatic workflow execution
+                                                                                  - 2 clarifying questions only
+Test cases:                                                                       - 0 failed API calls
+
+• Valid outputs generated                                                         - 6,000 tokens consumed
+
+• API calls succeed
+
+• Error handling works
+
+• Edge cases covered                                                           Using the skill-creator skill
+
+Example:                                                                       The skill-creator skill - available in Claude.ai via plugin directory or
+                                                                               download for Claude Code - can help you build and iterate on skills. If you
+                                                                               have an MCP server and know your top 2–3 workflows, you can build and test a
+   Test: Create project with 5 tasks                                           functional skill in a single sitting - often in 15–30 minutes.
+   Given: Project name "Q4 Planning", 5 task descriptions
+   When: Skill executes workflow                                               Creating skills:
+   Then:
+                                                                               • Generate skills from natural language descriptions
+      - Project created in ProjectHub
+      - 5 tasks created with correct properties                                • Produce properly formatted SKILL.md with frontmatter
+      - All tasks linked to project
+                                                                               • Suggest trigger phrases and structure
+      - No API errors
+
+                                                                               Reviewing skills:
+
+                                                                               • Flag common issues (vague descriptions, missing triggers, structural
+3. Performance comparison                                                        problems)
+
+                                                                               • Identify potential over/under-triggering risks
+Goal: Prove the skill improves results vs. baseline.
+                                                                               • Suggest test cases based on the skill's stated purpose
+Use the metrics from Define Success Criteria. Here's what a comparison might
+look like.                                                                     Iterative improvement:
+
+                                                                               • After using your skill and encountering edge cases or failures, bring those
+Baseline comparison:
+                                                                                 examples back to skill-creator
+
+                                                                               • Example: "Use the issues & solution identified in this chat to improve how the
+   Without skill:
+   - User provides instructions each time
+                                                                                 skill handles [specific edge case]"
+   - 15 back-and-forth messages
+   - 3 failed API calls requiring retry
+   - 12,000 tokens consumed
+
+                                                                                                                                                                  16
+To use:                                                                         Execution issues:
+
+                                                                                • Inconsistent results
+   "Use the skill-creator skill to help me build a skill for
+                                                                                • API call failures
+   [your use case]"
+                                                                                • User corrections needed
+
+
+Note: skill-creator helps you design and refine skills but does not execute
+                                                                                   Solution: Improve instructions, add error handling
+automated test suites or produce quantitative evaluation results.
+
+
+Iteration based on feedback
+Skills are living documents. Plan to iterate based on:
+
+
+Undertriggering signals:
+
+• Skill doesn't load when it should
+
+• Users manually enabling it
+
+• Support questions about when to use it
+
+
+
+   Solution: Add more detail and nuance to the description - this may include
+   keywords particularly for technical terms
+
+
+
+Overtriggering signals:
+
+• Skill loads for irrelevant queries
+
+• Users disabling it
+
+• Confusion about purpose
+
+
+
+   Solution: Add negative triggers, be more specific
+
+
+
+
+                                                                                                                                        17
+Chapter 4
+
+
+
+
+Distribution and
+sharing            18
+  Chapter 4
+
+
+
+Distribution and sharing
+Skills make your MCP integration more complete. As users compare connectors,         Using skills via API
+those with skills offer a faster path to value, giving you an edge over MCP-only
+alternatives.                                                                        For programmatic use cases - such as building applications, agents, or automated
+                                                                                     workflows that leverage skills - the API provides direct control over skill
+                                                                                     management and execution.
+Current distribution model (January 2026)
+                                                                                     Key capabilities:
+How individual users get skills:
+                                                                                     •    `/v1/skills` endpoint for listing and managing skills
+1. Download the skill folder
+                                                                                     • Add skills to Messages API requests via the `container.skills` parameter
+2. Zip the folder (if needed)
+                                                                                     • Version control and management through the Claude Console
+3. Upload to Claude.ai via Settings > Capabilities > Skills
+                                                                                     • Works with the Claude Agent SDK for building custom agents
+4. Or place in Claude Code skills directory
+
+                                                                                     When to use skills via the API vs. Claude.ai:
+Organization-level skills:
+• Admins can deploy skills workspace-wide (shipped December 18, 2025)                    Use Case                                           Best Surface
+• Automatic updates
+                                                                                         End users interacting with skills directly         Claude.ai / Claude Code
+• Centralized management
+                                                                                         Manual testing and iteration during development    Claude.ai / Claude Code
+
+An open standard
+                                                                                         Individual, ad-hoc workflows                       Claude.ai / Claude Code
+We've published Agent Skills as an open standard. Like MCP, we believe skills
+should be portable across tools and platforms - the same skill should work               Applications using skills programmatically         API
+whether you're using Claude or other AI platforms. That said, some skills are
+designed to take full advantage of a specific platform's capabilities; authors can       Production deployments at scale                    API
+
+note this in the skill's compatibility field. We've been collaborating with
+members of the ecosystem on the standard, and we're excited by early adoption.           Automated pipelines and agent systems              API
+
+
+
+
+                                                                                                                                                                      19
+Note: Skills in the API require the Code Execution Tool beta, which provides the
+secure environment skills need to run.                                                      - Select the skill folder (zipped)
+
+
+For implementation details, see:                                                        3. Enable the skill:
+                                                                                            - Toggle on the [Your Service] skill
+• Skills API Quickstart                                                                     - Ensure your MCP server is connected
+• Create Custom skills
+                                                                                        4. Test:
+• Skills in the Agent SDK                                                                   - Ask Claude: "Set up a new project in [Your Service]"
+
+
+
+Recommended approach today                                                           Positioning your skill
+Start by hosting your skill on GitHub with a public repo, clear README (for          How you describe your skill determines whether users understand its value and
+human visitors — this is separate from your skill folder, which should not contain   actually try it. When writing about your skill—in your README, documentation,
+a README.md), and example usage with screenshots. Then add a section                 or marketing - keep these principles in mind.
+to your MCP documentation that links to the skill, explains why using both
+together is valuable, and provides a quick-start guide.                              Focus on outcomes, not features:
+
+1. Host on GitHub                                                                    ✅ Good:
+  – Public repo for open-source skills
+  – Clear README with installation instructions                                         "The ProjectHub skill enables teams to set up complete project
+  – Example usage and screenshots                                                       workspaces in seconds — including pages, databases, and
+                                                                                        templates — instead of spending 30 minutes on manual setup."
+2. Document in Your MCP Repo
+  – Link to skills from MCP documentation
+  – Explain the value of using both together
+  – Provide quick-start guide
+                                                                                     ❌ Bad:
+3. Create an Installation Guide                                                         "The ProjectHub skill is a folder containing YAML frontmatter
+                                                                                        and Markdown instructions that calls our MCP server tools."
+
+   ## Installing the [Your Service] skill
+
+   1. Download the skill:                                                            Highlight the MCP + skills story:
+       - Clone repo: `git clone https://github.com/yourcompany/
+         skills`
+       - Or download ZIP from Releases                                                  "Our MCP server gives Claude access to your Linear projects.
+                                                                                        Our skills teach Claude your team's sprint planning workflow.
+   2. Install in Claude:                                                                Together, they enable AI-powered project management."
+       - Open Claude.ai > Settings > skills
+       - Click "Upload skill"
+                                                                                                                                                                 20
+Chapter 5
+
+
+
+
+Patterns and
+troubleshooting   21
+  Chapter 5
+
+
+
+Patterns and troubleshooting
+These patterns emerged from skills created by early adopters and internal teams.    Pattern 1: Sequential workflow orchestration
+They represent common approaches we've seen work well, not prescriptive
+templates.                                                                          Use when: Your users need multi-step processes in a specific order.
+
+                                                                                    Example structure:
+Choosing your approach: Problem-first vs. tool-first
+
+Think of it like Home Depot. You might walk in with a problem - "I need to fix a       ## Workflow: Onboard New Customer
+kitchen cabinet" - and an employee points you to the right tools. Or you might
+pick out a new drill and ask how to use it for your specific job.                      ### Step 1: Create Account
+                                                                                       Call MCP tool: `create_customer`
+Skills work the same way:                                                              Parameters: name, email, company
+
+• Problem-first: "I need to set up a project workspace" → Your skill orchestrates
+                                                                                       ### Step 2: Setup Payment
+  the right MCP calls in the right sequence. Users describe outcomes; the skill
+                                                                                       Call MCP tool: `setup_payment_method`
+  handles the tools.
+                                                                                       Wait for: payment method verification
+• Tool-first: "I have Notion MCP connected" → Your skill teaches Claude the
+  optimal workflows and best practices. Users have access; the skill provides          ### Step 3: Create Subscription
+  expertise.                                                                           Call MCP tool: `create_subscription`
+                                                                                       Parameters: plan_id, customer_id (from Step 1)
+Most skills lean one direction. Knowing which framing fits your use case helps
+you choose the right pattern below.                                                    ### Step 4: Send Welcome Email
+                                                                                       Call MCP tool: `send_email`
+                                                                                       Template: welcome_email_template
+
+
+
+
+                                                                                    Key techniques:
+
+                                                                                    • Explicit step ordering
+
+                                                                                    • Dependencies between steps
+
+                                                                                    • Validation at each stage
+
+                                                                                    • Rollback instructions for failures
+                                                                                                                                                          22
+Pattern 2: Multi-MCP coordination               Pattern 3: Iterative refinement
+Use when: Workflows span multiple services.     Use when: Output quality improves with iteration.
+
+Example: Design-to-development handoff          Example: Report generation
+
+
+   ### Phase 1: Design Export (Figma MCP)          ## Iterative Report Creation
+   1. Export design assets from Figma
+   2. Generate design specifications               ### Initial Draft
+   3. Create asset manifest                        1. Fetch data via MCP
+                                                   2. Generate first draft report
+   ### Phase 2: Asset Storage (Drive MCP)          3. Save to temporary file
+   1. Create project folder in Drive
+   2. Upload all assets                            ### Quality Check
+   3. Generate shareable links                     1. Run validation script: `scripts/check_report.py`
+                                                   2. Identify issues:
+   ### Phase 3: Task Creation (Linear MCP)             - Missing sections
+   1. Create development tasks                         - Inconsistent formatting
+   2. Attach asset links to tasks                      - Data validation errors
+   3. Assign to engineering team
+                                                   ### Refinement Loop
+   ### Phase 4: Notification (Slack MCP)           1. Address each identified issue
+   1. Post handoff summary to #engineering         2. Regenerate affected sections
+   2. Include asset links and task references      3. Re-validate
+                                                   4. Repeat until quality threshold met
+
+
+Key techniques:                                    ### Finalization
+                                                   1. Apply final formatting
+• Clear phase separation
+                                                   2. Generate summary
+• Data passing between MCPs                        3. Save final version
+
+• Validation before moving to next phase
+
+• Centralized error handling                    Key techniques:
+
+                                                • Explicit quality criteria
+
+                                                • Iterative improvement
+
+                                                • Validation scripts
+
+                                                • Know when to stop iterating
+
+                                                                                                         23
+Pattern 4: Context-aware tool selection                         Pattern 5: Domain-specific intelligence
+Use when: Same outcome, different tools depending on context.   Use when: Your skill adds specialized knowledge beyond tool access.
+
+Example: File storage                                           Example: Financial compliance
+
+
+   ## Smart File Storage                                           ## Payment Processing with Compliance
+
+
+   ### Decision Tree                                               ### Before Processing (Compliance Check)
+   1. Check file type and size                                     1. Fetch transaction details via MCP
+   2. Determine best storage location:                             2. Apply compliance rules:
+       - Large files (>10MB): Use cloud storage MCP                   - Check sanctions lists
+       - Collaborative docs: Use Notion/Docs MCP                      - Verify jurisdiction allowances
+       - Code files: Use GitHub MCP                                   - Assess risk level
+       - Temporary files: Use local storage                        3. Document compliance decision
+
+
+   ### Execute Storage                                             ### Processing
+   Based on decision:                                              IF compliance passed:
+   - Call appropriate MCP tool                                       - Call payment processing MCP tool
+   - Apply service-specific metadata                                 - Apply appropriate fraud checks
+   - Generate access link                                            - Process transaction
+                                                                   ELSE:
+   ### Provide Context to User                                       - Flag for review
+   Explain why that storage was chosen                               - Create compliance case
+
+
+                                                                   ### Audit Trail
+Key techniques:                                                    - Log all compliance checks
+                                                                   - Record processing decisions
+• Clear decision criteria
+                                                                   - Generate audit report
+• Fallback options
+
+• Transparency about choices
+                                                                Key techniques:
+
+                                                                • Domain expertise embedded in logic
+
+                                                                • Compliance before action
+
+                                                                • Comprehensive documentation
+
+                                                                • Clear governance
+
+
+                                                                                                                                      24
+Troubleshooting
+                                                         # Wrong
+                                                         name: My Cool Skill
+Skill won't upload
+
+Error: "Could not find SKILL.md in uploaded folder"      # Correct
+                                                         name: my-cool-skill
+Cause: File not named exactly SKILL.md
+
+Solution:
+                                                      Skill doesn't trigger
+• Rename to SKILL.md (case-sensitive)
+
+• Verify with: ls -la should show SKILL.md            Symptom: Skill never loads automatically
+
+Error: "Invalid frontmatter"                          Fix:
+
+Cause: YAML formatting issue                          Revise your description field. See The Description Field for good/bad examples.
+
+Common mistakes:                                      Quick checklist:
+
+                                                      • Is it too generic? ("Helps with projects" won't work)
+   # Wrong - missing delimiters                       • Does it include trigger phrases users would actually say?
+   name: my-skill
+   description: Does things
+                                                      • Does it mention relevant file types if applicable?
+
+                                                      Debugging approach:
+   # Wrong - unclosed quotes
+   name: my-skill
+                                                      Ask Claude: "When would you use the [skill name] skill?" Claude will quote the
+   description: "Does things
+                                                      description back. Adjust based on what's missing.
+
+   # Correct
+   ---
+                                                      Skill triggers too often
+   name: my-skill
+                                                      Symptom: Skill loads for unrelated queries
+   description: Does things
+   ---                                                Solutions:
+
+                                                      1. Add negative triggers
+Error: "Invalid skill name"
+                                                         description: Advanced data analysis for CSV files. Use for
+Cause: Name has spaces or capitals                       statistical modeling, regression, clustering. Do NOT use for
+                                                         simple data exploration (use data-viz skill instead).
+
+
+
+                                                                                                                                       25
+2. Be more specific                                                 Instructions not followed
+
+                                                                    Symptom: Skill loads but Claude doesn't follow instructions
+   # Too broad
+   description: Processes documents
+                                                                    Common causes:
+
+   # More specific                                                  1. Instructions too verbose
+   description: Processes PDF legal documents for contract review     – Keep instructions concise
+                                                                      – Use bullet points and numbered lists
+                                                                      – Move detailed reference to separate files
+3. Clarify scope
+                                                                    2. Instructions buried
+                                                                      – Put critical instructions at the top
+   description: PayFlow payment processing for e-commerce. Use        – Use ## Important or ## Critical headers
+   specifically for online payment workflows, not for general
+                                                                      – Repeat key points if needed
+   financial queries.
+                                                                    3. Ambiguous language
+
+
+                                                                       # Bad
+MCP connection issues                                                  Make sure to validate things properly
+
+Symptom: Skill loads but MCP calls fail
+                                                                       # Good
+                                                                       CRITICAL: Before calling create_project, verify:
+Checklist:
+                                                                       - Project name is non-empty
+1. Verify MCP server is connected                                      - At least one team member assigned
+  – Claude.ai: Settings > Extensions > [Your Service]                  - Start date is not in the past
+  – Should show "Connected" status
+
+2. Check authentication                                             Advanced technique: For critical validations, consider bundling a script
+  – API keys valid and not expired                                  that performs the checks programmatically rather than relying on language
+  – Proper permissions/scopes granted                               instructions. Code is deterministic; language interpretation isn't. See the Office
+  – OAuth tokens refreshed                                          skills for examples of this pattern.
+
+3. Test MCP independently                                           4. Model "laziness" Add explicit encouragement:
+  – Ask Claude to call MCP directly (without skill)
+  – "Use [Service] MCP to fetch my projects"                           ## Performance Notes
+  – If this fails, issue is MCP not skill                              - Take your time to do this thoroughly
+                                                                       - Quality is more important than speed
+4. Verify tool names
+                                                                       - Do not skip validation steps
+  – Skill references correct MCP tool names
+  – Check MCP server documentation
+                                                                    Note: Adding this to user prompts is more effective than in SKILL.md
+  – Tool names are case-sensitive
+                                                                                                                                                         26
+Large context issues
+
+Symptom: Skill seems slow or responses degraded
+
+Causes:
+
+• Skill content too large
+
+• Too many skills enabled simultaneously
+
+• All content loaded instead of progressive disclosure
+
+Solutions:
+
+1. Optimize SKILL.md size
+  – Move detailed docs to references/
+  – Link to references instead of inline
+  – Keep SKILL.md under 5,000 words
+
+2. Reduce enabled skills
+  – Evaluate if you have more than 20 - 50 skills enabled simultaneously
+  – Recommend selective enablement
+  – Consider skill "packs" for related capabilities
+
+
+
+
+                                                                           27
+Chapter 6
+
+
+
+
+Resources and
+references      28
+  Chapter 6
+
+
+
+Resources and references
+If you're building your first skill, start with the Best Practices Guide, then   Tools and Utilities
+reference the API docs as needed.
+                                                                                 skill-creator skill:
+Official Documentation                                                           • Built into Claude.ai and available for Claude Code
+
+                                                                                 • Can generate skills from descriptions
+Anthropic Resources:
+                                                                                 • Reviews and provides recommendations
+• Best Practices Guide
+                                                                                 • Use: "Help me build a skill using skill-creator"
+• Skills Documentation
+
+• API Reference                                                                  Validation:
+• MCP Documentation                                                              • skill-creator can assess your skills
+
+                                                                                 • Ask: "Review this skill and suggest improvements"
+Blog Posts:
+• Introducing Agent Skills
+                                                                                 Getting Support
+• Engineering Blog: Equipping Agents for the Real World
+
+• Skills Explained                                                               For Technical Questions:
+
+• How to Create Skills for Claude                                                • General questions: Community forums at the Claude Developers Discord
+
+• Building Skills for Claude Code
+                                                                                 For Bug Reports:
+• Improving Frontend Design through Skills
+                                                                                 • GitHub Issues: anthropics/skills/issues
+
+                                                                                 • Include: Skill name, error message, steps to reproduce
+Example skills
+
+Public skills repository:
+• GitHub: anthropics/skills
+
+• Contains Anthropic-created skills you can customize
+
+                                                                                                                                                          29
+                                                                                     Before upload
+Reference A: Quick                                                                     Tested triggering on obvious tasks
+
+
+checklist                                                                              Tested triggering on paraphrased requests
+
+                                                                                       Verified doesn't trigger on unrelated topics
+Use this checklist to validate your skill before and after upload. If you want
+                                                                                       Functional tests pass
+a faster start, use the skill-creator skill to generate your first draft, then run
+through this list to make sure you haven't missed anything.                            Tool integration works (if applicable)
+
+                                                                                       Compressed as .zip file
+
+Before you start
+                                                                                     After upload
+    Identified 2-3 concrete use cases
+
+    Tools identified (built-in or MCP)                                                 Test in real conversations
+
+    Reviewed this guide and example skills                                             Monitor for under/over-triggering
+
+    Planned folder structure                                                           Collect user feedback
+
+                                                                                       Iterate on description and instructions
+
+During development                                                                     Update version in metadata
+
+    Folder named in kebab-case
+
+    SKILL.md file exists (exact spelling)
+
+    YAML frontmatter has --- delimiters
+
+    name field: kebab-case, no spaces, no capitals
+
+    description includes WHAT and WHEN
+
+    No XML tags (< >) anywhere
+
+    Instructions are clear and actionable
+
+    Error handling included
+
+    Examples provided
+
+    References clearly linked
+
+
+
+
+                                                                                                                                      30
+                                                                     Security notes
+Reference B: YAML
+frontmatter                                                          Allowed:
+                                                                     • Any standard YAML types (strings, numbers, booleans, lists, objects)
+
+                                                                     • Custom metadata fields
+Required fields
+                                                                     • Long descriptions (up to 1024 characters)
+
+  ---
+                                                                     Forbidden:
+  name: skill-name-in-kebab-case
+  description: What it does and when to use it. Include specific     • XML angle brackets (< >) - security restriction
+  trigger phrases.
+                                                                     • Code execution in YAML (uses safe YAML parsing)
+  ---
+                                                                     • Skills named with "claude" or "anthropic" prefix (reserved)
+
+
+
+
+All optional fields
+
+  name: skill-name
+  description: [required description]
+  license: MIT # Optional: License for open-source
+  allowed-tools: "Bash(python:*) Bash(npm:*) WebFetch" # Optional:
+  Restrict tool access
+  metadata: # Optional: Custom fields
+    author: Company Name
+    version: 1.0.0
+    mcp-server: server-name
+    category: productivity
+    tags: [project-management, automation]
+    documentation: https://example.com/docs
+    support: support@example.com
+
+
+
+
+                                                                                                                                              31
+Reference C: Complete skill
+examples
+For full, production-ready skills demonstrating the patterns in this guide:
+
+• Document Skills - PDF, DOCX, PPTX, XLSX creation
+
+• Example Skills - Various workflow patterns
+
+• Partner Skills Directory - View skills from various partners such as Asana,
+  Atlassian, Canva, Figma, Sentry, Zapier, and more
+
+These repositories stay up-to-date and include additional examples beyond
+what's covered here. Clone them, modify them for your use case, and use them as
+templates.
+
+
+
+
+                                                                                  32
+claude.ai