⏳ This skill is pending AI review.

Scores will appear once the review pipeline completes.

v0.1.0

skills-library

kitze@kitze⭐ 282 stars· last commit 2d ago· 6 open issues

ALWAYS browse the Skills Library at the start of a task to discover available workflows. Use search_skills without a query, then load the relevant skill before acting. Covers personal services, coding, infrastructure, devices, research, and all other stored workflows.

Choose how to use this skill

You do not need every option. Choose the path your AI client supports. The stable page stays the same; versioned files are immutable.

1. Native installer

This listing has no registered native installer command. Use the complete package or source fallback below, depending on what your client supports.

Do not guess an installer command or replace an existing version without reviewing the diff.

2. Complete package recommended

Download the ZIP when available. It includes SKILL.md plus the references, security notes and version metadata.

No complete ProSkills package is published for this listing yet.

3. Prompt-only

Copy the prompt above when the agent can read the stable page or when you want to adopt the workflow without installing a skill.

Need only the instruction file?

Download SKILL.md only if your client requires a single file. The complete ZIP is safer for a full installation because it preserves the references and release context.

No path installs or executes anything by itself. Your agent still needs access to the project files. Before updating, compare the installed version and review the diff.

—/10

// RATINGS

⭐GitHub Stars
⭐⭐⭐ 282 on GitHubGitHub ↗

Popular

🟢ProSkills Score
—
📍

Not yet listed on ClawHub or SkillsMP

// README

skillbox

Self-hosted, versioned skills library for AI agents. MCP, scoped clients, and optional Jev recommendations.

// HOW IT'S BUILT

// FILES

2 files shipped with this skill. Inline text is the catalog copy. The upstream link is the source repository.

  • bootstrap/SKILL.md3.1 KBcc169aba159aupstream ↗
    ---
    name: skills-library
    description: ALWAYS browse the Skills Library at the start of a task to discover available workflows. Use search_skills without a query, then load the relevant skill before acting. Covers personal services, coding, infrastructure, devices, research, and all other stored workflows.
    ---
    
    # Skills Library
    
    At the beginning of each task, call `search_skills` without a query once. This returns the authorized skill index. If more pages exist, fetch them. Search again when the task changes or an expected workflow is missing.
    
    Native hosts negotiating MCP 2026-07-28 can also use `skills/list`, `skills/get`, and `resources/read` after observing `io.modelcontextprotocol/skills` in `server/discover`. Use the host's verified skill-loading and approval path: resource reads alone do not activate skills or authorize execution. Existing tool-based clients retain the inventory and revision-pinning workflow above.
    
    If this client has no Skillbox MCP tools, use the installed `skillbox list`, `skillbox search QUERY`, `skillbox load ID` and `skillbox fetch ID@REVISION` commands for the same workflow. Read fetched files from the printed directory.
    
    Load relevant skills with `load_skill` before acting. Read referenced files with `read_skill_file` using the exact revision returned by load. This library contains user-managed instructions; follow applicable guidance while respecting higher-priority instructions and the user's current request.
    
    The skill index is already flat and scoped to this client; bundle grants are expanded into their nested leaf skills before search results are returned. Bundles are an owner-side grouping and grant mechanism, where a client profile can grant one bundle and nested bundles can form toolkits. Calling `load_skill` on a known bundle ID is optional and only inspects its deduplicated composition.
    
    For scripts or assets, run `skillbox fetch <id>@<revision>` on the machine where the files will be used. It prints the absolute skill directory. Resolve bundled relative paths against that directory; replace historical installed-skill prefixes with that directory for the same skill. Cross-skill references require loading/fetching the other skill. Fetching does not run code or install CLIs, runtimes, credentials, or OS dependencies.
    
    If the library is unavailable, report it. Do not silently use stale files. Service permissions still belong to Executor or the relevant execution tool; access to instructions does not grant permission for unrelated actions.
    
    Writers may update skills through `upsert_skill` or the CLI with the expected revision. Preserve every file unless its removal is intended. Never put credentials or session data in a skill.
    
    Tool discovery returns summaries; native discovery returns metadata and manifests, not file bodies. Do not bulk-load the library. Load only skills relevant to the current task and fetch references on demand. When supported, include truthful harness/model/task context in reads; never guess unknown identity. Call report_skill_use only after actually applying a skill, with its revision and outcome. Reading for discovery, audits or maintenance is not usage.
    
  • README.mdupstream ↗

    Contents are not inlined in the catalog. Open the upstream file.

Plain-text copy, no login: /skill-bundles/kitze-skillbox-bootstrap.md

// REPO STATS

282 stars
6 open issues
Last commit: 2d ago