skill-writer_skill

This skill helps you design well-structured Agent Skills, draft SKILL.md, and validate frontmatter for IDE integrations.
  • Kotlin

2.5k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill tencentblueking/bk-ci --skill skill-writer

  • SKILL.md9.9 KB

Overview

This skill guides you to design and write well-structured Agent Skills for IDEs like CodeBuddy and Cursor. It covers scope definition, placement choices, manifest frontmatter, content organization, testing, and debugging. The goal is a focused, discoverable Skill that reliably activates for the right user queries.

How this skill works

I walk you through clarifying the Skill purpose, selecting project- or user-level locations, creating the directory layout, and authoring a manifest with required frontmatter fields. I provide concrete examples, validation checklists, testing steps, and debugging tips so the Skill activates predictably and follows best practices. I can also suggest names, create the main skill file content, and generate supporting files on request.

When to use it

  • You need to create a new Agent Skill for an IDE (project- or user-scoped).
  • You want help writing the skill manifest (frontmatter) and a clear description so the Skill is discoverable.
  • You need to design the Skill directory structure and optional support files (reference, scripts, config).
  • You want to convert prompts, workflows, or tools into a focused Skill.
  • You need to validate, test, or debug why a Skill does not activate.

Best practices

  • One Skill = one clear capability; keep scope narrow and actionable.
  • Name rules: lowercase, digits, hyphens only, match directory name, ≤64 chars.
  • Write a concrete description: what it does + when to use it + trigger phrases and file types.
  • Provide step-by-step instructions tailored for the AI assistant, not generic prose.
  • Use progressive disclosure: put advanced docs and examples in separate reference files.
  • Validate manifest syntax, file paths, and activation by testing real queries.

Example use cases

  • Create a pipeline-manager Skill that starts builds, queries status, and lists history when users mention CI/CD or pipeline.
  • Author a backend-development Skill that prescribes project layout, DI patterns, and testing guides for new microservices.
  • Convert a manual workflow (e.g., filling PDF forms) into a focused Skill with examples and helper scripts.
  • Build a tool-integration Skill that documents required IDs, parameters, and sample commands for invoking external CI tools.
  • Debug a Skill that never activates by tightening trigger phrases, verifying manifest fields, and confirming file placement.

FAQ

Project-level for team-shared workflows and version control; user-level for personal experiments and workflow variants.

What makes a description trigger reliably?

Include specific actions, file types, and common user phrases plus an explicit "when to use" clause so the assistant can match intent.

What validation steps should I run before sharing?

Confirm manifest syntax, directory name matches the manifest name, test activation with real queries, and check that examples are runnable.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
skill-writer skill by tencentblueking/bk-ci | VeilStrat