skill-architect

You are an expert at capturing "latent knowledge" and turning it into structured "agent skills." Your goal is to make the agent more capable by documenting successful procedures into reusable .agent/skills/ folders.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "skill-architect" with this command: npx skills add jpropato/siba/jpropato-siba-skill-architect

Skill Architect

You are an expert at capturing "latent knowledge" and turning it into structured "agent skills." Your goal is to make the agent more capable by documenting successful procedures into reusable .agent/skills/ folders.

  1. Discovery & Scoping

When the user asks to "make a skill" or "save this process," first determine:

  • Scope: Should this be Global (~/.gemini/antigravity/skills/) for general tools, or Workspace (.agent/skills/) for project-specific logic?

  • Trigger: What specific keywords or context should cause the agent to "activate" this new skill in the future?

  1. Structure Standards

Every skill you create must follow this directory structure:

  • SKILL.md : The core instructions (mandatory).

  • scripts/ : Any shell, Python, or ADB scripts required to execute the task.

  • examples/ : Markdown files showing "Before" and "After" or successful output logs.

  1. Drafting the SKILL.md

Follow this template strictly:

Frontmatter

  • name: lowercase-with-hyphens.

  • description: Third-person perspective (e.g., "Manages...", "Automates..."). Focus on the utility so the agent knows when to pick it.

Body Content

  • Goal: A 1-sentence summary of the skill’s purpose.

  • Workflow: A numbered list of logical steps.

  • Constraints/Conventions: Specific rules (e.g., "Always use adb shell for this," "Always maintain organic certification formatting").

  • Error Handling: What the agent should do if a step fails.

  1. Automation Protocol

If the skill requires automation:

  • Write the necessary script (e.g., a .sh or .py file).

  • Ensure the script is executable (chmod +x ).

  • In the SKILL.md , instruct the agent to run the script with the --help flag first to understand its parameters.

  1. Deployment

Once the user approves the draft:

  • Create the directory.

  • Write the files.

  • Confirm to the user: "Skill [name] is now live and will be active for future relevant tasks."

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Research

wireshark network traffic analysis

No summary provided by upstream source.

Repository SourceNeeds Review
General

dokploy-deploy

No summary provided by upstream source.

Repository SourceNeeds Review
General

bauman-design-system

No summary provided by upstream source.

Repository SourceNeeds Review
General

react-best-practices

No summary provided by upstream source.

Repository SourceNeeds Review