name: skill-writer description: >- When the user requests to create, update, or refine a GitHub Copilot skill, generate or revise a complete SKILL.md file that strictly adheres to the official format, validation rules, and community best practices.

Skill Writer

Generate or update SKILL.md files for GitHub Copilot coding agents, ensuring precise activation, concise expert-level guidance, and full compliance with repository standards.

When to Activate

•User explicitly asks to create or update a skill for a specific issue, topic or task.
•User requests to update, improve, or refactor an existing SKILL.md.
•User provides new rules, examples, or feedback intended to guide skill authoring.
•Capture Agent Insights: When the agent struggles with a recurring task, encounters repeated failures, or discovers an effective new solution/workaround not already documented, update the relevant existing skill (or create a new one) to incorporate the insight, ensuring future activations handle the case more efficiently.

Core Process

•
Infer a unique, descriptive name in lowercase-hyphenated format that will exactly match the intended folder name.
•
Write a single, highly precise description sentence that captures exact activation triggers without overlap.
•
Validate that the YAML frontmatter follows all repository-defined schema, linting, and validation rules (e.g., required fields, no extra keys, correct formatting).
•
Structure the file exactly as:
- •First line: 
- •YAML frontmatter block
- •Pure Markdown body with minimal, value-adding sections only
•
Write dense, imperative, expert-level instructions assuming ninja proficiency; skip basics, favor one-liners, pack maximum depth.
•
Design for fully autonomous agent execution: include self-contained loops, error recovery, verification steps; avoid any user confirmation except safety-critical cases.
•
Keep total length to minimum.
•
Section Discipline: Use sections sparingly and only when they clearly improve readability or organization; always follow a logical progression (e.g., Brief Intro → When to Activate → Core Process → Diagnostics → Patterns → What to Avoid → Limitations → Safety).
•
Output only the complete, ready-to-commit SKILL.md content.

Guidance on Creating or Updating Skills

When creating or updating a skill file, ensure the following format:

•
YAML Frontmatter is Mandatory: Begin every SKILL.md with required fields name (lowercase-hyphenated, exactly matching folder name) and description (one precise, activation-triggering sentence explaining exactly what the skill does and when Copilot should load it). Include license only if needed. YAML must pass all repository validation and linting rules.
•
Start with Lint Disable: Always open the file with  to suppress common markdownlint warnings.
•
Description Precision: Write the description as a single, highly specific sentence that matches user intent patterns without overlap—poor phrasing causes missed or false activations.
•
Pure Markdown Body: Use only Markdown in the body; never include extraneous files, scripts, or resources unless explicitly required for the skill.
•
Section Discipline: Use sections sparingly and only when they add clear value; prioritize logical flow (e.g., Core Process → Diagnostics → Patterns → Limitations → Safety).
•
Autonomous Execution Focus: Write for non-interactive agent flow: self-contained loops, built-in error recovery, verification steps, and no user confirmation prompts except for safety-critical actions.
•
Skill Length Management: If a skill exceeds or approaches 400 lines, or if content becomes unfocused, split it into multiple specialized skills with narrow, non-overlapping activation descriptions to maintain conciseness, precise triggering, and efficient token usage.
•
Avoid Hardcoding: Never embed specific values, file paths, repository names, user details, or tool versions; instead, use clear placeholders (e.g., <repository-name>, <file-path>, <version>).
•
Command-Line Tool Examples: When the skill involves command-line tools, derive them dynamically via tools/read operations to maximize skill generality and cross-project reusability. Always provide useful examples of key commands—especially for complex, multi-flag, piped, or non-standard operations—to ensure precise reproduction, minimize agent misinterpretation, and reduce execution failures.